Hedgehog-Gli Signaling in Human Disease (Molecular Biology Intelligence Unit)

MOLECULAR BIOLOGY INTELLIGENCE UNIT Hedgehog-Gli Signaling in Human Disease Ariel Ruiz i Altaba, Ph.D. Department of ...

Author: Ariel Ruiz i Altalba (Editor)

45 downloads 1263 Views 18MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

MOLECULAR BIOLOGY INTELLIGENCE UNIT

Hedgehog-Gli Signaling in Human Disease

Ariel Ruiz i Altaba, Ph.D. Department of Genetic Medicine and Development University of Geneva Medical School Geneva, Switzerland

LANDES BIOSCIENCE /EUREKAH.COM GEORGETOWN, TEXAS

U.S.A.

SPRINGER SCIENCE+BUSINESS MEDIA NEW YORK, NEW YORK

U.SA

HEDGEHOG-GLI SIGNALING IN HUMAN DISEASE Molecular Biology Intelligence Unit Landes Bioscience / Eurekah.com Springer Science+Business Media, Inc. ISBN: 0-387-25784-5

Printed on acid-free paper.

Copyright ©2006 Eurekah.com and Springer Science+Business Media, Inc. All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher, except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. T h e use in the publication of trade names, trademarks, service marks and similar terms even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the authors, editors and publisher believe that drug selection and dosage and the specifications and u s ^ e of equipment and devices, as set forth in this book, are in accord with current recommendations and practice at the time of publication, they make no warranty, expressed or implied, with respect to material described in this book. In view of the ongoing research, equipment development, changes in governmental regulations and the rapid accumulation of information relating to the biomedical sciences, the reader is urged to carefiiUy review and evaluate the information provided herein. Springer Science+Business Media, Inc., 233 Spring Street, New York, New York 10013, U.S.A. http://www.springer.com Please address all inquiries to the Publishers: Landes Bioscience / Eurekah.com, 810 South Church Street, Georgetown, Texas 78626, U.S.A. Phone: 512/ 863 7762; FAX: 512/ 863 0081 http://www.eurekah.com http: //www. landesbioscience. com Printed in the United States of America. 9 8 7 6 5 4 3 2 1 Cover art: 'Genomes III' ©Ariel Ruiz i Altaba (www.ruizialtaba.com). Courtesy of the artist.

Library of Congress Cataloging-in-Publication Data Ruiz i Altaba, Ariel. Hedgehog-gli signaling in human disease / Ariel Ruiz i Altaba. p. ; cm. ~ (Molecular biology intelligence unit) Includes bibliographical references and index. ISBN 0-387-25784-5 1. Cellular signal transduction. 2. Oncogenes. 3. T u m o r proteins. I. Title. II. Series: Molecular biology intelligence unit (Unnumbered) [DNLM: 1. Signal Transduction—physiology. 2. Oncogene Proteins—metabolism. 3. Transcription Factors—metabolism. Q U 375 R934h 2006] QP517.C45R85 2006 616.99'4071-dc22

2005028604

Dedication To Robert Gorlin for the formulation of the syndrome that carries his name and for his story of a Gorlin s syndrome patient carrying a snake in a taxi that crashed, who was bitten by the snake and had a broken leg, as told at a recent meeting. To all the past and present members of the Ruiz i Altaba lab, including Nadia Dahmane, Barbara Stecca, Virginie Clement, Veronica Palma, Pilar Sanchez, Christophe Mas and Jose Mullor. To our collaborators, to the scientists in this exciting field, to the contributors to this book, and to the patients who give their tissues and cells for research. I am gratefiil to Barbara Stecca, Virginie Clement, Mare Zbinden, Simon Restrepo and Christophe Mas for help in the editing of this volume.

CONTENTS Preface 1. How the Hedgehog Outfoxed the Crab: Interference with HEDGEHOG-GLI Signaling as Anti-Cancer Therapy? Ariel Ruiz i Altaha 2. The Patched Receptor: Switching On/OflFthe Hedgehog Signaling Pathway Luis Quijaddy Ainhoa Callejoy Carlos Torroja and Isabel Guerrero The Hedgehog Signaling Pathway Sequence and Functional Analysis of Ptc Hedgehog Lipid Modifications and Morphogen Distribution Reception of H h H h Internalization and Signal Transduction Smo Regulation by Ptc Ptc and Human Disease 3. Making a Morphogenetic Gradient Henk Roelink Displ, A Molecule Dedicated to the Export of Shh Is Shh Handed Over from Cell to Cell Using an Active Mechanism? 4. Spatial and Temporal Regulation of Hair Follicle Progenitors by Hedgehog Signaling Anthony E. Oro The Hair Follicle Cycle InterfoUicular Epithelial Cells Are Competent to Induce Shh Target Genes Outcomes of Shh Target Gene Induction in the Skin Regulation of Shh Induction in a Multipotent Epithelium Regulation of Shh Signal Reception Maintenance of Shh Signaling during Tissue Morphogenesis Future Perspectives 5. Mode of PTCHl/Ptchl-Associated Tumor Formation: Insights from Mutant Ptchl Mice Heidi Hahn Is Loss of Bodi PTCHl/Ptchl Alleles Required for Tumor Formation? What Are the Modifiers o( PTCHl/Ptchl'Associated Tumorigenesis?

xiii

1

23 23 25 25 26 26 27 28 34 35 38

41 43 43 AG 47 47 49 49

53

55 58

6. Basal Cell Carcinomas, Hedgehog Signaling, and die Ptchl+Z- Mouse Ervin Epstein, Jr. Rhabdomyosarcomas MeduUoblastomas BCCs 7. GLI Genes and Their Targets in Epidermal Development and Disease Fritz Aherger and Anna-Maria Frischauf GLI Genes and Epidermal Development Human Epidermis and the Origin of BCC GLIl and GLI2 in Basal Cell Carcinoma The Oncogenic Nature of GLI Transcription Factors Regulation of GLI Expression Unresolved Problems and Future Perspectives 8. Splitting Hairs: Dissecting the Roles of Gli Activator and Repressor Functions during Epidermal Development and Disease Pleasantine Mill and Chi-Chung Hui The GH Family of Transcription Factors in H h Signaling Molecular Mechanisms of Embryonic and Adult Hair Development Shh: Stem Cells, Cell Cycle and Cancer 9. Shh Expression in Pulmonary Injury and Disease PaulM. Fitchy SoniaJ. Wakelin, Jacqueline A. Lowreyy William A.H. Wallace and Sarah EM. Howie The Hedgehog Pathway Hedgehog Signalling in Normal Pulmonary Tissues Hedgehog Signalling in Injury and Disease Shh in the Immune System Hedgehogs and Pulmonary Progenitor Cells Shh and Type II Epithelial Cells 10. Human Correlates of GLI3 Function Leslie G. Biesecker The Delineation of the Greig Cephalopolysyndactyly Syndrome (GCPS) The Molecular Basis of GCPS The Clinical Delineation of Pallister-Hall Syndrome (PHS) Animal Models of GLI3 Phenotypes

63 65 65 G(i

74 75 77 80 80 82 82

86 86 95 101 119

119 120 120 123 124 125 129

129 130 130 131

11.

From Oligodactyly to Polydactyly: Role of Shh and Gli3 in Limb Morphogenesis Chin Chiang Shh Expression and the Specification of Digit Identity Congenital Limb Malformations Associated with Shh Misregulation Shh Activity in Limb Patterning Is Mediated by Gli3 Congenital Limb Malformations Are Associated with Gli3 Mutations

137 138 139 I4l 142

12. The Genetics of Indian Hedgehog M. Elizabeth McCready and Dennis E. Bulman Brachydactyly Type Al Acrocapitofemoral Dysplasia (ACFD) Comparison of Human I H H Disorders I H H Animal Model

146

13. Sonic Hedgehog Signaling in Craniofacial Development Dwight Cordero, Minal Tapadia and Jill A. Helms The Face in Myth and Reality The Craniofacial Consequences of Disrupting S H H Signaling: The Human Experience Models to Study Craniofacial Development Craniofacial Consequences of Shh Perturbation: The Model Experience Embryologic Aspects of Normal Craniofacial Development: An Introduction Future Directions

153

14. Important Role of Shh Controlling Gli3 Functions during the Dorsal-Ventral Patterning of the Telencephalon Jun Motoyama and Kazushi Aoto Shh Signaling and Dorsal-Ventral Patterning of the Telencephalon Gli3 as an Activator of Dorsal Cell Types Gli3 as a Suppressor of Ventral Cell Type Importance of the Balance between Gli3 and Shh Future Perspectives 15. Regtdation of Early Events in Cell Cycle Progression by Hedgehog Signaling in CNS Development and Tumorigenesis Anna Marie Kenney and David H. Rowitch Mitogenic Signaling in the Developing CNS S H H Signaling Is Required for Growth of Cerebellar Granule Neuron Precursors and Is Etiologic in MeduUoblastoma N-MYC as a Key Component of H H Signaling

147 149 149 150

153 156 158 158 161 171

177

177 182 182 183 183

187 187 193 197

16. Modulating the Hedgehog Pathway in Diseases Frederic J. de Sauvage and Lee L. Rubin Identification of Small Molecule Hedgehog (Hh) Antagonists Identification of Hh Small Molecule H h Agonists

210

17. Hedgehog Signaling in Endodermally Derived Tumors Marina Pasca di Magliano and Matthias Hehrok The Hedgehog Signaling Pathway Hedgehog Signaling in Gastrointestinal Tumors of the Pancreas and Colon

215

Index

211 212

215 218 225

,

CFMT'rÔ

1

.

jQr.

Ariel Ruiz i Altaba Department of Genetic Medicine and Development University of Geneva Medical School Geneva, Switzerland Email: [email protected] Chapter 1

U-^—-—^-—^ rÔlMTT^ TT5T inrrô c = v>WiM i

Fritz Aberger Department of Molecular Biology Division of Genomics University of Salzburg Salzburg, Austria Email: fritz. [email protected]. at Chapter 7 Kazushi Aoto Molecular Neuropathology Group Brain Science Institute, RIKEN Wako, Saitama, Japan Chapter 14 Leslie G. Biesecker National Human Genome Research Institute National Institutes of Health Bethesda, Maryland, U.S.A. Email: [email protected] Chapter 10 Dennis E. Bulman Ottawa Health Research Institute Department of Biochemistry, Microbiology and Immunology and Department of Medicine Division of Neurology University of Ottawa Ottawa, Ontario, Canada Email: [email protected] Chapter 12

Ainhoa Callejo Centro de Biologia Molecular "Severo Ochoa" Universidad Autonoma de Madrid Cantoblanco, Madrid, Spain Chapter 2 Chin Chiang Department of Cell and Developmental Biology Vanderbilt University Medical Center Nashville, Tennessee, U.S.A. Email: [email protected] Chapter 11 Dwight Cordero Department of Obstetrics and Gynecology Brigham and Women's Hospital and Harvard Medical School Boston, Massachusetts, U.S.A. Chapter 13 Frederic J. de Sauvage Department of Molecular Biology Genentech, Inc. South San Francisco, California, U.S.A. Email: [email protected] Chapter 16 Ervin Epstein, Jr. Department of Dermatology University of California, San Francisco San Francisco, California, U.S.A. Email: [email protected] Chapter 6

|

Paul M. Fitch Immunobiology Group MRC Centre for Inflammation Research College of Medicine and Veterinary Medicine University of Edinburgh Edinburgh, U.K. Chapter 9 Anna-Maria Frischauf Department of Molecular Biology Division of Genomics University of Salzburg Salzburg, Austria Chapter 7 Isabel Guerrero Centro de Biologia Molecular "Severo Ochoa" Universidad Aut6noma de Madrid Cantoblanco, Madrid, Spain Chapter 2 Heidi Hahn Institute of Human Genetics University of Gottingen Gottingen, Germany Email: [email protected] Chapter 5 Matthias Hebrok Diabetes Center Department of Medicine University of California, San Francisco San Francisco, California, U.S.A. Email: [email protected] edu Chapter 17 Jill A. Helms Department of Plastic and Reconstructive Surgery Stanford University Stanford, California, U.S.A. Email: [email protected] Chapter 13

1

Sarah E.M. Howie Immunobiology Group MRC Centre for Inflammation Research College of Medicine and Veterinary Medicine University of Edinburgh Edinburgh, U.K. Email: [email protected] Chapter 9 Chi-Chung Hui Program in Developmental Biology The Hospital for Sick Children and Department of Molecular and Medical Genetics University of Toronto Toronto, Ontario, Canada Email: [email protected] Chapter 8 Anna Marie Kenney Department of Pediatric Oncology Dana-Farber Cancer Institute Boston, Massachusetts, U.S.A. Chapter 15 Jacqueline A. Lowrey Immunobiology Group MRC Centre for Inflammation Research College of Medicine and Veterinary Medicine University of Edinburgh Edinburgh, U.K. Chapter 9 M. Elizabeth McCready Ottawa Health Research Institute Ottawa, Ontario, Canada Chapter 12 Pleasantine Mill Comp. and Developmental Genetics MRC Human Genetics Unit Western General Hospital Edinburgh, U.K. Email: [email protected] Chapter 8

Jun Motoyama Molecular Neuropathology Group Brain Science Institute, RIKEN Wako, Saitama, Japan Email: [email protected] Chapter 14 Anthony E. Oro Program in Epithelial Biology Stanford University Stanford, California, U.S.A. Email: [email protected] Chapter 4 Marina Pasca di Magliano Diabetes Center Department of Medicine University of California, San Francisco San Francisco, California, U.S.A. Chapter 17 Luis Quijada Centro de Biologia Molecular "Severo Ochoa" Universidad Autonoma de Madrid Cantoblanco, Madrid, Spain Chapter 2 Henk Roelink Department of Biological Structure University of Washington Seatde, Washington, U.S.A. Email: [email protected] Chapter 3 David H. Rowitch Department of Pediatric Oncology Dana-Farber Cancer Institute Boston, Massachusetts, U.S.A. Email: [email protected] Chapter 15

Lee L. Rubin Curis, Inc. Cambridge, Massachusetts, U.S.A. Chapter 16 Minal Tapadia Department of Plastic and Reconstructive Surgery Stanford University Stanford, California, U.S.A. Chapter 13 Carlos Torroja The Wellcome Trust Sanger Institute The Wellcome Trust Genome Campus Hinxton, Cambridge, U.K. Chapter 2 Sonia J. Wakelin Immunobiology Group MRC Centre for Inflammation Research College of Medicine and Veterinary Medicine University of Edinburgh Edinburgh, U.K. Chapter 9 William A.H. Wallace Immunobiology Group MRC Centre for Inflammation Research College of Medicine and Veterinary Medicine University of Edinburgh Edinburgh, U.K. Chapter 9

PREFACE

T

he last two decades have seen an explosion in the understanding of how specific mechanisms of cellular communication regulate embryonic development, adult homeostasis, stem cell behavior and cancer. One of these systems that cells use to talk to each other and thereby know their position utilizes Hedgehog glycoproteins. Hedgehog signals regulate key transcription factors of the Gli family in the responding cells. The Hedgehog-Gli pathway is used many times in many organs in the life of an individual, including a number of stem cells and derived progenitors. Insufficient signaling in the embryo can lead to cyclopia and other birth defects while enhanced function can lead to tumor formation. Why? What is the context and rationale? What do the Gli proteins regulate? What is the exact link between embryonic development, stem cell lineages and cancer? Can Hedgehog-Gli signaling be used to regulate stem cell lineages in regenerative medicine? These are some of the questions that the chapters in this primer address. A decade has passed since the discovery of the Hedgehog genes in vertebrates and the establishment of causal links between Hedgehog signaling and cyclopia, Gorlin's syndrome and familial basal cell carcinomas by a number of laboratories. An important step in the understanding of human sporadic cancer came four years ago with the finding that sustained signaling is required for the proliferation of brain tumors, a result from our laboratory that has since been expanded to an increasing number of cancers. Thus, we are now witnessing the birth of a great new hope for the understanding of cancer and other diseases that is taking place hand in hand with the elucidation of the molecular mechanisms orchestrating embryonic development. It is with this enthusiasm that the present book has been edited, knowing ftill well that all scientific knowledge is, per necessity, partial and in need of revision and expansion, and that great hopes and expectations are ofi:en just that. However, considering the enormous advances made after the discovery of Hedgehog-Gli function in animal development and cancer described in the chapters in this book, it is possible to suggest that we are just entering a long ascending ramp, one that opens new avenues for further scientific exploration. Importantly, the research in this book also raises possibilities for the development of rational, anti-cancer therapies targeting Hedgehog-GLI signaling in the very near future. Ariel Ruiz i Altaba Geneva, Switzerland 2005

CHAPTER 1

How the Hedgehog Outfoxed the Crab: Interference with HEDGEHOG-GLI Signaling as Anti-Cancer Therapy? Ariel Ruiz i Altaba

T

he enormous success of the last two decades in molecular embryology and developmental genetics have afforded fresh views on old problems. For instance, from a developmental point of view there is an alternative way to think about solid cancers that contrast with the single cell focus derived from molecular and cell biology and their emphasis on the cell cycle and the concepts of transformation and oncogene. One can consider cancers as diseases of patterning affecting stem cell lineages or properties. Within this framework, the funnel hypothesis suggests that the multiple ways to induce a given type of tumor in the mouse and the many functional mutations identified in a given tumor in humans lead to the activation of a single or few events in the cell that affect its position and identity. A one to one relationship between phenotype and final molecular mechanism with some redundancy is proposed for cancer, much as in embryonic development. Moreover, the ability of pathologists to recognize consistent morphologies in tumors and to give appropriate diagnoses is explained by the action of consistent paradevelopmental programs.' Concepts like cell competence, specification, niche, developmental history and positional information thus become central to understand, and hopefully treat, cancer. Indeed, under this light, understanding cancer becomes necessary in order to have a complete view of developmental potential. We can then consider—with caution as we presently lack the perspective that elapsed time affords—^what has been learned and achieved on Hedgehog signaling in cancer in recent years. The Hedgehog story is an unfolding one, one that does not yet have a known end, and is only one of several stories,^ but so far it is the most promising one. This story starts with the discovery of the Hedgehog (Hh) gene in Drosophila melanogaster -25 years ago, well before Hh signaling was implicated in cancer. The identification of the Hh gene through its mutation, which causes aberrant embryonic patterning giving the mutant larva a 'hairy' aspect, and thus named hedgehog,^ was published in 1980. The cloning of the single Drosophila Hh gene was published by several groups in the early '90s, ' with the cloning and identification o^Hh genes in vertebrates {Sonic hedgehog (Shh), Indian Hh and Desert Hh in mice and humans) following shortly afterwards. The definition of the expression patterns of Hh genes in the developing embryo showed that they are often strikingly restricted to cells and areas previously known to have special patterning activities. The exquisite spatio-temporal regulation ofHh genes in the developing embryo and adult tissues, together with the ability of Hh proteins to elicit defined cell fate changes in a concentration-dependent manner highlighted a revolution in molecular embryology, providing mechanistic explanations for critical events in *Ariel Ruiz i Altaba—Department of Genetic Medicine and Development, University of Geneva Medical School, 8242 CMU, 1 rue Michel Servet, CH-1211 Geneva 1204, Switzerland. Email: [email protected]

Hedgehog-Gli Signaling in Human Disease^ edited by Ariel Ruiz i Altaba. ©2006 Eurekah.com and Springer Science+Business Media.


Therapeutic

targets Cydopamine, SANT1-4

A .first

Cur61414, Hh-Antag691

+/- Modulators B:best GLI"®p peptides Antisense RNAi

Figure 1. Scheme of the HEDGEHOG-GLI pathway. The action of HH, PTCHl and SMOH at the membrane level is emphasized, together with the site ofaction of natural and artificial inhibitors. The former include HIP, GAS and PTCHl itself, while the latter include cyclopamine and anti-HH antibodies. The action of SMOH leads to the activity of the GLI proteins. Resident GLI proteins in the cytoplasm, likely bound to the cytoskeleton are activated and translocate into the nucleus where they regulate target gene transcription. The function of the GLI proteins depends on the function of several modifiers acting downstream of SMOH. The scheme emphasizes the action of endogenous modifiers as well as artificial blockers of GLI function, including GLI repressor peptides, antisense molecules and RNAi. Inhibitors to the upper component of the pathway (A) have lead the way, but inhibitors of the lower part, inhibiting directly or indirectly GLI function (B), are predicted to be superior as these will inhibit pathway fiinction by any activating event at any level. the formation of the animal body plan. This includes for instance, the patterning and growth of limb buds and of the brain and spinal cord.^'^^ T h e activity of the H h pathway involves several steps (Fig. l).^"^'^ Secreted H h glycoproteins act via the transmembrane proteins Patchedl (Ptchl) and Smoothened (Smo). It is thought that H h binding to Ptchl inhibits the repression of Smo by Ptchl so that Smo functions only when H h is present. Smo in turn transduces the H h signal intracellularly to activate the zinc-finger Gli transcription factors. ^^ T h e Gli proteins are required and sufficient to mediate H h activity. A number of modulators, positive or negative, some involving negative feedback such H i p l , Gasl and Ptchl itself, also regulate Gli activity so that this is tightly controlled in time and space. Because the H h - G l i pathway is to a large extent linear, activity can be achieved by loss of function of negative factors, such as Ptchl or the gain of function of positive factors, such as mutations in Smo that may render it insensitive to Ptchl repression. T h e initial link of HH signaling a n d familial h u m a n cancer was published in the mid-1990s^5,i6 jinking mutations in PATCHED 1 {PTCHl, an inhibitor of H H signaling; Fig. 1), with the rare Basal Cell Nevus syndrome or Gorlin's syndrome.^^ Affected individuals can develop numerous basal cell carcinomas (BCCs) of the skin at an early age as well as meduUoblastomas and other tumors, and display additional malformations including odontogenic keratocysts.^^ These studies converged on PTCHl from the study of its function in flies and mammals and from skin carcinogenesis.^^ T h e first demonstration of a functional link of H H - G L I signaling with sporadic h u m a n cancer was published in 1997, proposing that all h u m a n B C C s — t h e most c o m m o n type of

How the Hedgehog Outfoxed the Crab

cancer—derive from the activation of the HH-GLI pathway as all examined BCCs expressed GLIl^^ (Fig. 1), since its transcription is the best marker of a cell's response to H H signaling. ^^ In addition to GLIly GLI2 and PTCHl (another H H regulated gene ^) have also been found to be expressed consistently in sporadic BCCs.^^'^^'"^^ GLIl was discovered as an amplified gene in a human glioma line in 1987. This was followed by the discovery of a related gene in vertebrates {GLI3^^) and independently of a single related gene, Cubitus interruptus (Ci), in Drosophila?^'^^ A role for GLIl in cancer was negated by subsequent fmdings^^' and research on GLIl waned for several years. However, GLI3 was shortly afterwards identified as the gene mutated in Greig's Cephalopolydactily syndrome^^" and later implicated in Pallister-Hall syndrome.^^'^ Patients affected with these diseases display a wide range of phenotypes, including additional digits, skull defects, imperforate anus and hypothalamic growths (hamartomas). The misregulation of post-transcriptionally cleaved C-terminally-deleted forms of GLI3 acting as dominant repressors, as first shown for Ci, has been implicated in these syndromic phenotypes with debated genotype-phenotype correlations. ^' ' Aberrant GLI function has also been implicated in other syndromic phenotypes, including VACTERL, ' a syndrome that includes vertebral anomalies, anal atresia, cardiac defects, tracheoesophageal fistula, and renal and limb defects. Three Gli genes were then cloned from different species, including mice, ^' ^ chicks,^^ frogs, ^^'^^ and zebrafish.^^'^^ They were originally shown to affect neural patterning by mimicking Shh.^^'^^ In addition, microinjection of human or frog Glil lead to sporadic epidermal and CNS tumors or hyperplasias (Fig. 2).^^'^^ Interpretation of GLIl-induced epidermal hyperplasias as tumors related to human BCCs awaited the papers on the involvement of loss oiPTCHl in Gorlins syndrome in which patients develop numerous BCCs. Following the finding of familial mutations in PTCHl in Gorlins syndrome patients^^' and simultaneous with the involvement of HH-GLI signaling in sporadic cancer,^ mutations in the HH-GLI pathway components PTCHl and SMOHifls^. 1) were sought and found in a small fraction of sporadic skin, brain and other tumors, ' ' some of which have been proven to have functional relevance. There are also reported mutations in Suppressor of fused (SUFUH; but see re£ 69), a negative modulator of GLI function.^^'^^ Mutations in the GLI genes have not been found in tumors, with the exception of beta-actin-GLIl fusions in a subset of pericytomas.^^ These findings provide possible bases for the activation of the H H pathway in a fraction of tumors although many others might derive from epigenetic events. Various animal models demonstrating the sufficiency of an active Hh-Gli pathway to induce tumor formation (Table 1, Fig. 2) began to be published in 1997 for skin (Shh'/"^ Glil;^^''^^''^^ Smo;^^'^^ Ptchl'J^ Glif'^^'^^-^^). brain {Ptchlf^^^^ GlilP Shh^c-Myc^^^ ShhÎGF2'^^^ Shh+AkP'^^) and muscle {Ptchl^^). At the same time, a rational context for the various human tumors now associated with HH-GLI signaling began to be uncovered with the realization that elements of this pathway are expressed in the corresponding normal tissues and that HH-GLI signaling controls growth, patterning and/or homeostasis of developing and adult organs such as the skin^^'^^'^'^'^ând the cerebellum^^'^^ (Fig. 3). Indeed, the control of cell proliferation in the late embryonic and perinatal dorsal brain by SHH-GLI signaling may account for its involvement in many types of CNS tumors.^^ Shh-Gli signaling is not only required for the normal expansion of many types of progenitors but it is also required for the control of neural stem cell behavior in the perinatal and adult rodent brain, including the embryonic neocortex^^ (Fig. 4), the perinatal and adult hippocampus^^'^^ and the subventricular zone of the lateral ventricle of the forebrain^^'^^ (Fig. 4). Shh also regulates the proliferation of embryonic spinal cord precursors.^^ Similarly, there is evidence that Hh signaling regulates precursor/stem cell lineages and morphogenesis in other tissues and organs, such as the intestine and perhaps the skin. ' Together, these findings have raised the possibility that tumors derive from HH-sensitive stem cell lineages, possibly acting at the level of stem cells in some cases or at the level of derived early precursors. Generally, the best evidence that stem cells exist in tumors come from the transfer of leukemia from a sick mouse to a healthy recipient by the implantation of a few rare cancer stem cells '


GLIl

GUI

GLIl+MOGlil

Figure 2. Induction of tumors by misexpressed GLI1. Misexpression of human G LI 1 in the developing frog embryo leads to CNS (A,C,D) and skin (B) tumors. Coinjection of GLIl RNA plus LacZKNA allows the identification of tumor-forming cells as descendants of the cells that received the injected materials and are blue after X-Gal reaction (A,C,D). Labeling of cycling cells with BrdU (B, green) shows tumor hyperproliferation. Coinjection of human GLII plus a morpholino-modified antisense oligonucleotide specific for the endogenous frog Glil RNA leads to tumor suppression (D), indicating the requirement of sustained endogenous Glil function for tumorigenesis. Reprinted with permission from refs. 18 and 55. and the impressive findings that teratocarcinoma cells can contribute to the development of an entire fertile mouse when transplanted into the inner cell mass of a recipient embryo. Here, cancer would appear to be induced by genetic/epigenetic changes that are reversed in the appropriate context or niche. These original experiments demonstrate the stemness of teratocarcinoma cancer cells and their totipotency. For adult solid tumors, evidence in favor of the existence of cancer stem-like cells derives from the finding that cells with self-renewal properties have been reported in tumor cells from brain and breast. ' Since a small number of glioma cells show clonogenic properties that are regulated by FiFi signaling (P. Sanchez, V. Clement and ARA, pers. com.) stem cell lineages (stem cells a n d / o r their early derived more restricted progeny) may thus be the target of carcinogens, mutations and epigenetic changes that lead to the initiation of tumorigenesis in the brain and other HH-responsive organs and tissues, such as the epithelial compartment of the prostate, lung or gastrointestinal tract. T h e first demonstration that h u m a n sporadic cancers require sustained FiH-GLI signaling for continued growth derived from the ability to inhibit the proliferation of meduUoblastoma cells by blocking H H signaling. Since then, a rapidly increasing number of sporadic h u m a n tumors has been found to express GLIl and sometimes 5 / / / / ( F i g . 5), and to be dependent on sustained HFi-GLI pathway function (Table 2). These include basal cell carcinomas; ' meduUoblastomas,^ '^^^ gUomas (R Sanchez, V. Clement and ARA, pers. com.), pancreatic


Table 1. Human cancers dependent on sustained HH-GLI signaling Tumor

Inhibitors

Target*

References 55

Cyclopamine

PC, CL

Cyclopamine

PC, MA, CL

112

Glioma

Cyclopamine

PC, CL

55

Medulloblastoma

Basal cell carcinoma

Cyclopamine

patients

169

Oral squamous cell carcinoma

Cyclopamine

CL

125

Small cell lung cancer

Cyclopamine Hh antibody

MA, CL

115

Pancreatic cancer

Cyclopamine

MA, CL

113

Hh antibody

MA, CL

116

Cyclopamine

CL

114

Digestive tract cancer


PC, MA, CL

116

Prostate cancer

Cyclopamine RNA interference Hh antibody

PC, CL

117


MA, CL

118

Cyclopamine

CL

19

Breast cancer

Cyclopamine

CL

121

Colorectal cancer

Cyclopamine

CL

129

* PC: primary culture; MA: mouse xenograft; CL: cell line

t u m o r s 113,114 , " ' " " ' small cell lung cancer, stomach tumors and prostate tumors. 117-120 Additional studies show that a variety of (benign and malignant) tumor cells express HH-GLI pathway components, that tumors are correlated with mutation in genes that affect the pathway and/or that tumor cell lines are sensitive to inhibition of H H signaling. These include breast tumors,^^^ sarcomas,^^^ ameloblastomas, ^^^ a subset of pericytomas,^^ astrocytomas, ' oral squamous cell carcinomas,^^^ bone exostoses,^ neurofibromas,^^^ bladder cancer,^^^ colorectal tumors^^^'^^"^ and plasmacytomas^^^ among others. The number of tumors and benign tumor-like growths, such as odontogenic keratocycts, in which HH-GLI signaling may play a critical role is presently growing at an impressive rate. For example, tumors of liver, colon, rectum, lung and stomach are reported to have low levels of Hedgehog-interacting protein^^^ (HIPl), a H H signaling antagonist.^^^ The rationale for the initial testing for the requirement of sustained HH-GLI function for tumor maintenance first derived from the timing of appearance of somatic GLIl-induced tumors in tadpoles^^'^^ and from the finding that the activity of mutations in SMOH and PTCHl is inhibited by cyclopamine,^^ although the latter cannot distinguish between an involvement in tumor initiation or tumor maintenance. The appearance of GLIl-induced tadpole tumors (or hyperplasias) occurred well after all the injected material was degraded, suggesting the presence of'memory'. As the endogenous Glil gene was found to be expressed in the induced tumors, it was hypothesized that the activation of the endogenous pathway leading to Glil expression was involved in tumor development and could account for the delay in tumor appearance. To test this idea, a morpholino-modified antisense oligonucleotide specific for the endogenous Glil mRNA (selectively inhibiting its translation) was coinjected


Maintained _ SHH-GLI signaling

BRAIN TUMOR oEGL iEGL

PL

GLl IGL Cerebellar cortex GUI Figure 3. SONIC HEDGEHOG signaling controls cerebellar patterning and growth and its deregulation leads to meduUoblastoma. Schematic representation of a section through the developing cerebellar cortex showing the layered disposition of cell types and the action of SHH from Purkinje neurons (pink) in the Purkinje layer (PL) regulating the proliferation of granule neurons precursors (green) in the outer external germinal layer (oEGL). These then become postmitotic (light blue), move inwards to the internal EGL (iEGL) and migrate to form the internal granule layer (deep blue, IGL) where they terminally differentiate displaying an axon and dendrites. SHH signaling is regulated to allow proliferation during a defined time frame to produce the correct number of precursors needed for a given species. Granule neuron progenitors in the EGL normally undergo a GLIl^ to GLIl' transition concomitant with differentiation and their inward migration (light blue arrow). Inappropriate maintenance of an active SHH-GLI pathway (red arrow) in precursors (red) leads to hyperproliferation of G L I P cells and tumor formation. The case of cerebellar medulloblastomas exemplify the derivation of tumors from stem cells/precursors that utilize the SHH-GLI pathway or proliferation, expansion or renewal. Reprinted with permission from ref 86. with synthetic human GLIl RNA. Coinjected embryos developed normally without tumors, demonstrating the requirement of the endogenous GUI protein for tumor development. This result prompted us to originally test the requirement of sustained signaling for tumor maintenance in humans, even if Glil was reported to be redundant in mice. ' T h e requirement of sustained signaling for tumor maintenance and growth has been also recently demonstrated in mice. Conditional expression of Gli2 in basal skin cells leads to the formation of BCCs, which regress when the expression is turned off after tumor formation. This result is in line with previous data on the dependence of mouse tumors on the maintained expression of the inducing oncogene. ^^^'^ ^ Inhibition of H H - G L I function in human tumor cells was first achieved with cyclopamine, is a plant alkaloid with an overall resemblance to cholesterol that is produced by Veratrum californicum, a corn lily.^ '^'^ ^ It was named cyclopamine because it is the active compound, along with jervine, responsible for the production of cyclopic embryos and newborns after ingestion of V. califomicum by pregnant range and farm animals.^^^'^^^'^^^ Cyclopamine-induced cyclopia phenocopies the loss ofShh in mice^ ^ and humans^ ^'^^^ and it selectively inhibits the H H - G L I pathway,^^^ binding and blocking the function of S M O H (Fig. 1).^^^ Inhibition of H H - G L I signaling by noncyclopamine small molecule agents ' that, like cyclopamine, target S M O H has been achieved in mouse basal cell carcinoma punch explants and cell lines.^^^' Antibodies against SHH^^ and immunization with H I P l have also been


A

Embryonic neocortex 16 E l 8.5

E18.S

14

25

% 12

S20 c o

c o u

-sis

en

c u «i QL

S 10 c u 4J q

6 4 2

-*-

0 wt

as

0

Shh-/-

wt

GII2 - / -

B wt

wt #

Shh-/-

•

GII2-/''*ii''*^^'*

c

#

Adult forebrain subventricular zone

35

500

Adult ^

30

450 400

T

* 25

/r/?i transcription continues in mid anagen as the hair epithelium increases in size. During this time ptchl staining is seen in dermal papilla cells and the outer root sheath (Fig. 2g). At the start of catagen when apoptosis occurs, expression of Shh and many of its target genes ceases. Expression is undetectable in quiescent (telogen) hairs or anywhere in the interfoUicular skin (Fig. 2d). The expression of Shh and the reception of signaling are tighdy controlled in time and space within the skin.

InterfoUicular Epithelial Cells Are Competent to Induce Shh Target Genes Shh signaling is normally restricted to the cells of the hair follicle/sebaceous gland lineage. Restricted signaling could occur because reception of the Shh signal is prevented in epidermal progenitor cells or merely because Shh is not expressed there. Important to understanding the


44

epideriiiial progenilors

S h h i i III ^ I y ^^^^ pfogcnitor cells BMP

epidermal progenitors

^#^«*«il^»F«*

# iI-r p"-

»pra^nttor ceils

t>iff»r«ntl»ticMi I

Early Atiageti Mlf// hair progenitors BMP

W»^»«»«J»"*

No|«tin

Mid Anagen

Shhl .

- I fi'

Figure 2. Shh signaling is spatially and temporally restricted, a) Diagram of the ability of the bulge progenitors to contribute to interfoUicular epidermal progenitors and hair follicle progenitors. Shh acts on hair follicle progenitors in the hair follicle niche to induce proliferation. From reference 75. b) in situ hybridization of anagen hair follicle showing the expression of Shh and c) ptchl at the tip of the growing hair epithelium, d) X-gal staining oiptchl-\2icL mice at telogen showing no Shh target gene induction in the resting hair, e) X-gal staining of early anagen hair epithelium showing Shh target induction in progenitors (blue) f) X-gal staining of early anagen hair as it grows out from previously resting hair. Much of the staining is also in the dermal papillae cells, g) X-gal staining of mid anagen hair showing Shh target gene induction in dermal papilla and outer root sheath staining. Reprinted with permission from ref 30.

spatial and Temporal Regulation of Hair Follicle Progenitors by Hedgehog Signaling

45

Figure 3. Many interfoUicular cells are capable of inducing Shh target genes, a) X-gal staining from Keratin l4-Glil;/>ft:Ai-lacZ mice at anagen shows most of the basal cells of the interfoUicular epidermis (arrows) express Shh target genes, b) Different section showing X-gal stained epithelial buds deriving from interfoUicular epithelium (arrows). Reprinted with permission from re£ 30. role of Shh signaling in disease processes is the effects of ectopic signaling in ectopic locations outside the hair follicle niche. Recent data suggests that interfoUicular cells are also capable of inducing canonical Shh targets such 2iSptchl 2J\dglil. The competence for epidermal progenitor cells to respond to Shh signaling was first suggested by experiments exposing interfoUicular epidermis to dermal papilla cells and isolated morphogens normally found within the hair follicle niche. Recombination experiments using hairless human foreskin and the dermal papillae cells within the niche resulted in sebaceous glands and rudimentary hair follicles, processes which require Shh signaling, while recombination with regular dermal fibroblasts yielded just interfoUicular epithelium. ' Further, forced expression in the interfoUicular cells of an activated |3-catenin, a integral component of the Wnt pathway normally expressed during anagen, induced ectopic hair foUicle morphogenesis in interfoUiciJar epithelium with concominant Shh and Shh target gene expression.^^'^^ Two additional experiments demonstrate that the epidermal progenitor cells are competent to direcdy induce Shh target genes. In 2A\Atptchll+, small tumor buds expressing Shh target gene induction arise from the interfoUicular epidermis (Fig. 3). While this suggests that there may be some cells that can respond to Shh signaling, it did not test which of the cells could respond. Using a transgenic promoter driving the glil gene in the background o(ptchl-IsicZy cells capable of responding to Glil could be visualized by X-gal staining. In anagen animals, the staining occurred throughout the epithelium, both foUiciJar and interfoUicular. This demonstrates that during anagen the entire basal epithelium appears to be capable of inducing Shh target genes if stimulated. These data fiirther suggest that the difference between cells destined


46

Outcomes of Shh target gene induction in the skin Loss of Shh target gene irtduction Epidermis

Hair Follicle

Gain of Shh target gene induction Epidermis

Minima! effects on growth and differentiation

Abnormal hair rrôrphogenesis, small with some hair diffentiation markers but lacking sebaceous glands

Hair follicle tumors such as basal cell carcinomas. trichoblastomas, hair follicle hamartomas

Example: Shh knockout Shh blocking antibody

Example: Shh knockout, Shh blocking antibody

Example: Keratin 14-Shh, Keratin 5 Gin, Gli2 Keratin 5-Smo transgenic mice

Hair Follicle Loss of differentiation markers Hair follicle timing

Example: Msx2-noggin,Shh adenovirus

Figure 4. Outcomes of Shh target gene induction depending on where expression occurs. to become hair follicle and stratified epidermis is not lineage, but the restricted exposure and/ or response to powerful morphogens from the niche.

Outcomes of Shh Target Gene Induction in the Skin While it is apparent that many cells are competent to induce Shh target genes, alterations in Shh signaling have different effects on the different compartments in which they occur and may be helpful in understanding pathway-dependent diseases (Fig. 4). Complete lack of Shh in the skin in Shh-null mice demonstrates the essential role in hair progenitor growth and morphogenesis. Shh mutant mice develop normally spaced hair placodes, which are precursors to mature hairs, but fail to imdergo normal hair morphogenesis. '^^ Differentiation markers demonstrate that much of normal hair differentiation occurs, but the outgrowth of the progenitor cells is blunted. Shh mutants also fail to express sebaceous glands, a differentiation pathway from the hair follicle, suggesting hedgehog signaling (likely involving cooperation from another hedgehog, Indian hedgehog) signaling is involved in sebaceous differentiation.^^'^^ Similarly, in adult hair follicle, antibodies that block Shh function arrest hair morphogenesis, indicating not a single requirement, but continued requirements for Shh function throughout anagen. 40 While epithelial Shh target gene induction regulates the growth of hair progenitor cells, the Shh pathway appears to have a more subtle requirement in the proliferation of progenitor cells that give rise to the stratified epithelium. Initial studies suggest that Shh is not required for differentiation of epidermal progenitor cells because shh mutants and animals injected with anti-Shh antibodies display a wild-type pattern of stratified epithelial differentiation.^^'^^'^^ Moreover, tumors of epithelial differentiation such as squamous cell carcinomas (SCCs) do not express high levels of the canonical Shh target gene ptchl. '"^^ However, Shh signaling may play a minor role in the growth of SCCs. SCCs are found at slightly higher frequency in irradiated/>^c/;i/+ mice, and mutations in the humsin ptchl gene are detected at low frequency in sporadic SCCs. Additionally, studies in human keratinocytes suggest Shh, in the absence ofglil induction, can oppose cell cycle arrest, ^ and restoration ofptchl function in an SCC cell lines can suppress cell growth in vitro. These data suggest that Shh signaling may contribute to differential cell growth during tumor formation. While loss of Shh target gene induction has litde effect on the development of epithelial progenitors, ectopic expresssion can have profound effects. Inappropriate activation of Shh target genes causes formation of hair follicle-derived tumors, of which the most clinically significant are BCCs. Beside BCC, a variety of other hair follicle and sebaceous tumors^^' have been shown to derive from Shh pathway activation. Many of these tumors have been shown to be associated with Shh target gene induction at interfoUicular locations. In transgenic animals overexpressing activating components of the Shh pathway in both the outer root sheath and


47

interfollicular epidermis, epithelial buds come from die interfoUicular epithelium.^^'^ Similar findings are seen in UV irradiated ptchll-v mice, with trichoblastomas and BCC-like lesions arise from epidermis not adjacent to a hair (Fig. 3b). ^ By contrast, the effects of increased Shh signaling in the context of the developing hair follicle appear to have quite different effêcts. Shh expressed from the bovine keratin 6 promoter that directs Shh expression to the inner root sheath failed to induce a distinct hair phenotype (A. Oro, unpublished). Furthermore, forced induction of Shh via transgenic mice expressing noggin in the hair is associated with abnormal differentiation, but lacks hair follicle tumors even in older mice. While additional experiments using hair-specific promoters may discover particular conditions that woidd suggest otherwise, the available data point to a restriction of the Shh effêcts on progenitor proliferation when it occurs within the context of the hair follicle niche.

Regulation of Shh Induction in a Multipotent Epithelium The ability of ectopic Shh target gene induction to redirect cell fates in cells outside the hair follicle underscores the critical need for spatial and temporal restriction of Shh target gene induction for controlled hair follicle growth and the prevention of follicular tumorigenesis. How then does the adult hair normally regulate Shh target gene induction in progenitor cells? The hair follicle niche, created by inductive signals during embryogenesis, continues to control the spatial and temporal expression of Shh in the adult. Key regulators of the developing niche include the FGFs, which promote Shh expression in the hair placode and the bone morphogenetic proteins (BMPs), which antagonize Shh expression. ^' ^ As in embryonic hair development, the domain of epithelial Shh expression in the postnatal hair appears to be limited by BMP activity in the differentiating hair progenitor cells. Noggin, an inhibitor of BMPs, is secreted from the underlying DP cells to antagonize BMP activity, allowing Shh expression to promote hair follicle growth. Loss of noggin fiinction results in increased BMP activity and the decreased number and size of hair follicles, while application of noggin-soaked beads can induce anagen and Shh expression.^^ Noggin overexpression in the hair matrix cells, using the Msx2 promoter, results in a broadening of Shh expression which is associated with an increased domain of proliferating hair follicle progenitor cells and decreased markers of differentiation. Recent results suggest that along with noggin to repress BMP, the induction of Wnt target genes is also required for Shh induction in the embryonic hair follicle niche. Canonical Wnt signaling requires the presence of a member of the TCF family of transcription factors and the persistence of its cofactor P-catenin. In the developing hair follicle niche, noggin expressed from the mesenchyme induces the expression of LEF-1, a founding member of the TCF transcription factor family. In turn Wnt expression stabilizes p-catenin expression allowing Wnt target gene induction to occur. With such signaling that includes the decreased expression of the adhesion molecule E-cadherin, the hair niche is formed and Shh expression occurs. Among the evidence that this occurs upstream of Shh expression is that E-cadherin repression occurs prior to Shh expression and occurs even in Shh mutants.

Regulation of Shh Signal Reception In addition to spatial and temporal regulation of Shh production in the niche, controlling Shh signal reception also plays an important role. Since many epidermal progenitor cells are capable of responding to Shh signaling, proper restriction of target gene induction acts to limit hair progenitor growth in a multipotent sheet of cells. The first example of such restriction was demonstrated in the developing Xenopus epithelium. ^^ Ectopic floor plate differentiation, a hallmark of Shh effects on the early nervous system, was not detected in response to injected Shh in the neural plate stage. With closure of the neural tube came floor plate differentiation, but it was still restricted predominandy to the dorsal region of the neural tube. Additional studies in chick and mouse confirm these original studies. Injection of Shh-expressing retroviruses into developing feather bud epithelium reveal a select developmental

48


Figure 5. Post-transcriptional regulation ofgli 1 during the hair cycle, a, b) Keratin-14-Gli 1 animals (TRANS) appears identical with wild-type (WT) littermates until anagen. With the phenotype comes accumulation of Gli 1 immunoreactivity, suggesting a post-transcriptional regulation during anagen. Reprinted with permission from ref 30. window where target gene and follicular bud formation is possible. In ptch -/+ mice, small ectopic tumor buds form in interfollicular epithelium as measured by X-gal staining coming from the/>/r/^/-lacZ mice.^^' Interestingly, the buds are only detected during anagen and not other stages of the hair cycle, suggesting that the epithelium is restricted to target gene induction during the stage when Shh signaling normally occurs. Additional experiments point to the regulation of the gli transcription factor as part of the basis for the signaling restriction. Transgenic ^//7^^'^ and gli2^^ animals are generally indistinguishable from their wild-type littermates at birth, a striking difference from those overexpressing Shh in the skin, which induces drastic morphological changes at the time it is expressed. Like xhcptchll+ mice, the Keratin 14-G*/// mice induced target genes and developed epithelial proliferation in interfollicular epithelium with the onset of anagen, supporting an anagen signal that activates glil. The activation is posttranscriptional in that glil UNA is present, but Gli protein is only detected during anagen (Fig. 5).^^ Fiow protein accumidation is regulated over the hair cycle is unknown, but is likely to occur at many levels, glil translation has been shown to be regulated through alternative 5' UTR splicing.^^ Moreover, stability of Gli family members have been linked to proteosome-dependent degradation via phosphorylation by protein kinase A, casein Kinase and glycogen synthase kinase p. A combination of alternative splicing and phosphorylation may contribute to reception of the Shh signal. An open question is the nature of the signal that regulates Gli accumulation. One possibility is that Shh itself, which is present at anagen, functions to regidate Gli. Shh is known to allow preformed Gli to become nuclear and could also be responsible for Gli accumulation. This hypothesis is appealing in that ectopic Shh can regulate the timing of anagen. Injection of Shh-expressing adenoviruses into mouse skin prematurely triggers the onset of anagen, ^ and thus can influence hair cycle staging. Fiowever, Shh is expressed in the invaginating matrix cells


49

a great distance from the interfoUicular epidermis (Fig. 2b) making it less likely to be able to influence the epidermal cells direcdy. Shh involvement would have to be part of an indirect or long range signaling mechanism.

Maintenance of Shh Signaling during Tissue Morphogenesis In order to precisely pattern developing tissues, morphogens must induce Shh ligand production and the presence of pathway members in receiving cells. However, additional regulators are required to both maintain Shh signaling once started and terminate it appropriately. In the developing limb, Shh expression and limb outgrowth is maintained via a positive feedback loop with FGF4 that is mediated by the BMP antagonist gremlin. '^^ Shh induces gremlin in the adjacent mesenchyme which relieves BMP-mediated repression of FGF in the overlying ectoderm. In the hair follicle, little is known about what maintains Shh signaling. The juxtaposition and functional relationship between Shh and the dermal papilla-derived BMP antagonist noggin (Fig. 2a) would suggest a similar positive feedback loop. Indeed, ectopic noggin expression in the hair matrix induces Shh levels and the presence of Shh target gene induction in noggin-expressing cells supports this hypothesis. However, expression of the activator form of Gli2 in the epithelium rescues the skin phenotype of a Gli2 mutant, suggesting that Shh signal reception is not necessary in the dermal papilla for hair morphogenesis. This argues that any Shh regulation of dermal papilla function must occur indirectly through additional, as yet unidentified factors. Candidate proteins include the Wnt family, which are known to be targets of Shh in the hair ' and function to maintain dermal papilla competence. Along with factors that maintain signaling, factors that precisely shut off signaling are important in tissue morphogenesis. In the hair follicle, proteins that regulate the anagen-to-catagen transition fit into this class. Perhaps the best studied example is Fibroblast growth factor 5 (FGF5), a protein expressed in the outer root sheath of growing hairs.^^ Mutations in FGF5 significantly extend the period of anagen and fail to shut off Shh expression. Homozyous angora mice lack FGF5 protein and have extremely long hair shafts, indicative of continued anagen. Interestingly, while Shh expression persists, angora mice do not get hair follicle tumors or BGCs, indicating that other factors are still present in the hair matrix to regulate the outcome of Shh signaling. Additional evidence comes from studies of plakoglobins effects on hair cycling. Plakoglobin is a p catenin-like protein that also binds to members of theTCF transcription factor family. In contrast to (i-catenin, which facilitates Shh induction, overexpression of plakoglobin represses epithelial growth, proliferation, and results in an early transition to catagen with the resultant shortened hair foUicles.^^ Given the known functions of (i-catenin in regulating the hair cycle, it is tempting to speculate that plakoglobin acts through competition with p-catenin, but the data is also consistent with P-catenin independent mechanisms.

Future Perspectives The precise delivery and reception of morphogenic signals such as those from Shh underlies the ability for the body to regenerate tissues as part of normal homeostasis or during wound repair. Summation of signals that regulate the initiation, reception and maintenance of these growth factors allow precise regeneration each time. This chapter attempts to separate different aspects of signaling to illustrate the importance of each one during morphogenesis. However, in reality a single factor is likely to contribute to one or more process. For example, Shh ligand expression is clearly dependent on Wnt signaling during early anagen, but P-catenin and members of the TCF transcription factors are also present in the receiving cells. Xgal staining using a Wnt target gene P-galactosidase readout indicates that Wnt signaling is also occurring, suggesting that Wnt may be regulating aspects of Shh reception and signal maintenance. Because current genetic techniques in mice or chick lack single cell resolution, the ability to remove or add gene function in donor or receiving cells at different times during hair follicle morphogenesis must be developed. Recent advances using

50


estrogen receptor-regulated ere recombinase^ could allow such temporal regulation and the identification of stage-specific promoters could provide means to address the issue. The data summarized in this chapter point out the continued need to study the differences in outcomes of cells exposed to powerful morphogenic signals like Shh both inside and outside the hair follicle niche. The present data suggests that many of the cells in the interfoUicular epidermis are capable of responding to the signals and that induction of Shh target genes in cells outside the hair follicle matrix where they normally exist can lead to ectopic hair follicle progenitor proliferation and follicular tumors. In particular, understanding the mechanisms of restricting Shh reception may be useful both for patterning and for the development of anti-tumor therapies. This will be a particularly exciting area of research given the growing number of tumors in which the Shh signal pathway has been implicated.

Acknowledgments I would Uke to thank M.P. Scott, P.A. Khavari, and members of the Oro lab for helpful suggestions. This work funded by NIH grants ROl AR046786 and POl AR44012.

References 1. 2. 3. 4.

Chuong C M . Molecular basis of epithelial appendage morphogenesis. Austin: RG Landes, 1998. Gumbiner B. Epithelial morphogenesis. Cell 1992; 69:385-387. Hogan B. Morphogenesis. Cell 1999; 96:225-233. Fuchs E, Tumbar T, Guasch G. Socializing with the neighbors: Stem cells and their niche. Cell 2004; 116:769-78. 5. Owens D , Watt F. Contribution of stem cells and differentiated cells to epidermal tumors. Nat Rev Cancer 2003; 3:444-51. 6. W a t t F, Hogan B. O u t of Eden: Stem cells and their niches. Science 2000; 287:1427-30. 7. Lum L, Beachy PA. The hedgehog response network: Sensors, switches, and routers. Science 2004; 304:1755-9. 8. McMahon AP. More surprises in the Hedgehog signaling pathway. Cell 2000; 100:185-188. 9. Pasca di Magliano M, Hebrok M. Hedgehog signalling in cancer formation and maintenance. Nat Rev Cancer 2003; 3:903-11. 10. Belloni E et al. I d e n t i f i c a t i o n of Sonic h e d g e h o g as a c a n d i d a t e gene responsible for holoprosencephaly. Nat Genet 1996; 14:353-6. 11. Roessler E et al. Loss-of-function mutations in the human GLI2 gene are associated with pituitary anomalies and holoprosencephaly-Uke features. Proc Natl Acad Sci USA 2003; 100:13424-9. 12. Kang S, Graham J, Olney A et al. Gli3 frameshift mutations cause autosomal dominant Pallister-hall syndrome. Nat Genet 1997; 15:266-8. 13. Vortkamp A, Gessler M, Grzeschik KH. GLI3 zinc-finger gene interrupted by translocations in Greig syndrome families. Nature 1991; 352:539-40. 14. H a h n H et al. Mutations of the human homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome. Cell 1996; 85:841-851. 15. Johnson RL et al. H u m a n homolog of patched, a candidate gene for the basal cell nevus syndrome. Science 1996; 272:1668-71. 16. Xie T, Spradling AC. Decapentaplegic is essential for the maintenance and division of germline stem cells in the Drosophila ovary. Cell 1998; 94:251-60. 17. Pietsch T et al. Medulloblastomas of the desmoplastic variant carry mutations of the human homologue of Drosophila patched. Cancer Res 1997; 57:2085-8. 18. Raffel C et al. Sporadic medulloblastomas contain P T C H mutations. Cancer Res 1997; 57:842-5. 19. Taylor M D et al. Mutations in SUFU predispose to medulloblastoma. Nat Genet 2002; 31:306-10. 20. H a h n H et al. Rhabdomyosarcomas and radiation hypersensitivity in a mouse model of Gorlin syndrome. Nat Med 1998; 4:619-22. 2 1 . Watkins D N et al. Hedgehog signalling within airway epithelial progenitors and in small-cell lung cancer. Nature 2003; 422:313-7. 22. Cotsarelis G, Sun T T , Lavker RM. Label-retaining cells reside in the bulge area of pilosebaceous unit: Implications for follicular stem cells, hair cycle, and skin carcinogenesis. Cell 1990; 61:1329-37. 2 3 . Tumbar T et al. Defining the epithelial stem cell niche in skin. Science 2004; 303:359-63. 24. Morris RJ et al. Capturing and profiling adult hair follicle stem cells. Nat Biotechnol 2004; 22:411-7. 25. Stenn KS, Paus R. Controls of hair follicle cycling. Physiol Rev 2001; 81:449-494. 26. Oshima H , Rochat A, Kedzia C et al. Morphogenesis and renewal of hair follicles from adult multipotent stem cells. Cell 2 0 0 1 ; 104:233-45.


51

27. Taylor G, Lehrer MS, Jensen PJ et al. Involvement of follicular stem cells in forming not only the follicle but also the epidermis. Cell 2000; 102:451-61. 28. Pinkus H . Anatomy and histology of skin. In: Graham JJW, Helwig E, eds. Dermal Pathology. Harper & Row, 1972. 29. Paus R, Cotsarelis G. T h e biology of hair follicles. N Engl J Med 1999; 341:491-7. 30. O r o AE, Higgins K M . Hair cycle regulation of Hedgehog signal reception. Dev Biol 2 0 0 3 ; 255:238-48. 3 1 . Ferraris C, Bernard BA, Dhouailly D . Adult epidermal keratinocytes are endowed with pilosebaceous forming abilities. Int J Dev Biol 1997; 41:491-8. 32. Prouty SM, Lawrence L, Stenn KS. Fibroblast-dependent induction of a murine skin lesion with similarity to human common blue nevus. Am J Pathol 1996; 148:1871-85. 33. Lo Celso C, Prowse D M , Watt FM. Transient activation of beta-catenin signalling in adult mouse epidermis is sufficient to induce new hair follicles but continuous activation is required to maintain hair follicle tumours. Development 2004; 131:1787-99. 34. Gat U, DasGupta R, Degenstein L et al. De Novo hair follicle morphogenesis and hair tumors in mice expressing a truncated beta-catenin in skin. Cell 1998; 95:605-14. 35. Van Mater D , KoUigs FT, Dlugosz AA et al. Transient activation of beta -catenin signaling in cutaneous keratinocytes is sufficient to trigger the active growth phase of the hair cycle in mice. Genes Dev 2003; 17:1219-24. 36. St-Jacques B et al. Sonic hedgehog signaling is essential for hair development. Curr Biol 1998; 8:1058-68. 37. Chiang C et al. Essential role for Sonic hedgehog during hair follicle morphogenesis. Dev Biol 1999; 205:1-9. 38. Allen M et al. Hedgehog signaling regulates sebaceous gland development. A m J Pathol 2003; 163:2173-8. 39. Niemann C et al. Indian hedgehog and beta-catenin signaling: Role in the sebaceous lineage of normal and neoplastic mammalian epidermis. Proc N a d Acad Sci USA 2003; 100(Suppl 1):11873-80. 40. Wang LC et al. Regular articles: Conditional disruption of hedgehog signaling pathway defines its critical role in hair development and regeneration. J Invest Dermatol 2000; 114:901-8. 4 1 . Eklund LK et al. Mutation analysis of the human homologue of Drosophila patched and the xeroderma pigmentosum complementation group A genes in squamous cell carcinomas of the skin. Mol Carcinog 1998; 21:87-92. 42. Aszterbaum M et al. Ultraviolet and ionizing radiation enhance the growth of BCCs and trichoblastomas in patched heterozygous knockout mice. N a t Med 1999; 5:1285-91. 4 3 . Fan H , Khavari PA. Sonic hedgehog opposes epithelial cell cycle arrest. J Cell Biol 1999; 147:71-6. 44. Koike C et al. Introduction of wild-type patched gene suppresses the oncogenic potential of human squamous cell carcinoma cell lines including A 4 3 1 . Oncogene 2002; 21:2670-8. 4 5 . Sheng H et al. Dissecting the oncogenic potential of Gli2: Deletion of an NH(2)-terminal fragment alters skin tumor phenotype. Cancer Res 2002; 62:5308-16. 46. Kulessa H , T u r k G, Hogan BL. Inhibition of Bmp signaling affects growth and differentiation in the anagen hair folUcle. Embo J 2000; 19:6664-74. 47. Widelitz RB, Jiang TX, Noveen A et al. FGF induces new feather buds from developing avian skin. J Invest Dermatol 1996; 107:797-803. 4 8 . Noramly S, Morgan BA. BMPs mediate lateral inhibition at successive stages in feather tract development. Development 1998; 125:3775-87. 49. Jung H S et al. Local inhibitory action of BMPs and their relationships with activators in feather formation: Implications for periodic patterning. Dev Biol 1998; 196:11-23. 50. Botchkarev VA et al. Noggin is a mesenchymally derived stimulator of hair-follicle induction. N a t Cell Biol 1999; 1:158-64. 5 1 . Botchkarev VA et al. Noggin is required for induction of the hair follicle growth phase in postnatal skin. Faseb J 2 0 0 1 ; 15:2205-14. 52. Jamora C, DasGupta R, Kocieniewski P et al. Links between signal transduction, transcription and adhesion in epithelial bud development. Nature 2003; 422:317-22. 53. Ruiz i Altaba A, Jessell T M , Roelink H . Restrictions to floor plate induction by hedgehog and winged-helix genes in the neural tube of frog embryos. Mol Cell Neurosci 1995; 6:106-21. 54. Morgan BA, Orkin R W , Noramly S et al. Stage-specific effects of sonic hedgehog expression in the epidermis. Dev Biol 1998; 201:1-12. 55. Mancuso M et al. Basal cell carcinoma and its development: Insights from radiation-induced tumors in Ptchl-deficient mice. Cancer Res 2004; 64:934-41. 56. Nilsson M et al. Induction of basal cell carcinomas and trichoepitheliomas in mice overexpressing Gli-1. Proc Natl Acad Science 2000; 97:3438-43.

52

Hedgehog-Gli

Signaling in Human

Disease

57. Grachtchouk M et al. Basal cell carcinomas in mice overexpressing Gli2 in skin. Nature genetics 2000; 24:216-17. 58. O r o AE et al. Basal cell carcinomas in mice overexpressing Sonic Hedgehog. Science 1997; 276:817-821. 59. W a n g X Q , Rothnagel JA. Post-transcriptional regulation of the glil oncogene by the expression of alternative 5' untranslated regions. J Biol Chem 2 0 0 1 ; 276:1311-6. 60. Chen C H et al. Nuclear trafficking of Cubitus interruptus in the transcriptional regulation of Hedgehog target gene expression. Cell 1999; 98:305-16. 6 1 . Price MA, Kalderon D . Proteolysis of the Hedgehog signaling effector Cubitus interruptus requires phosphorylation by glycogen synthase kinase 3 and casein kinase 1. Cell 2002; 108:823-35. 62. W a n g Q T , Holmgren RA. T h e subcellular localization and activity of Drosophila cubitus interruptus are regulated at multiple levels. Development 1999; 126:5097-106. 6 3 . Sato N , Leopold PL, Crystal RG. Effect of adenovirus-mediated expression of sonic hedgehog gene on hair regrowth in mice with chemotherapy-induced alopecia. J N a d Cancer Inst 2001; 93:1858-64. GA. Capdevila J, Tsukui T, Rodriquez Esteban C et al. Control of vertebrate limb outgrowth by the proximal factor Meis2 and distalantagonism of BMPs by GremHn. Mol Cell 1999; 4:839-49. 6 5 . Zuniga A, Haramis AP, McMahon AP et al. Signal relay by BMP antagonism controls the S H H / FGF4 feedback loop in vertebrate limb buds. Nature 1999; 401:598-602. GG. Mill P et al. Sonic hedgehog-dependent activation of Gli2 is essential for embryonic hair follicle development. Genes Dev 2003; 17:282-94. G7. MuUor JL, Dahmane N , Sun T T et al. W n t signals are targets and mediators of Gli function. Curr Biol 2 0 0 1 ; 11:769-73. 68. Reddy ST et al. Characterization of W n t gene expression in developing and postnatal hair follicles and identification of W n t 5 a as a target of Sonic hedgehog in hair follicle morphogenesis. Mech Dev 2 0 0 1 ; 107:69-82. 69. Kishimoto J, Burgeson R, Morgan B. W n t signaling maintains the hair-inducing activity of the dermal papilla. Genes Dev 2000; 14:1181-5. 70. Hebert J M , Rosenquist T , Gotz J et al. FGF5 as a regulator of the hair growth cycle: Evidence from targeted and spontaneous mutations. Cell 1994; 78:1017-25. 7 1 . Charpentier E, Lavker R M , Acquista E et al. Plakoglobin suppresses epithelial proliferation and hair growth in vivo. J Cell Biol 2000; 149:503-20. 72. Maeda O et al. Plakoglobin (gamma-catenin) has TCF/LEF family-dependent transcriptional activity in beta-catenin-deficient cell Hne. Oncogene 2004; 23:964-72. 7 3 . DasGupta R, Fuchs E. Multiple roles for activated LEF/TCF transcription complexes during hair follicle development and differentiation. Development 1999; 126:4557-68. 7A. Metzger D , Clifford J, Chiba H et al. Conditional site-specific recombination in mammalian cells using a ligand-dependent chimeric Cre recombinase. Proc Natl Acad Sci USA 1995; 92:6991-5. 7 5 . Callahan CA, O r o AE. Monstrous attempts at adnexogenesis: Regulating hair follicle progenitors through Sonic hedgehog signaling. Curr O p i n Genet Dev 2 0 0 1 ; 11:541-6.

CHAPTER 5

Mode of PTCHl/Ptchl'Assod2Xtd Tumor Formation: Insights from Mutant Ptchl Mice Heidi Hahn* Abstract 'T^tchedl {PTCHl/Ptchl) is a member of the SHH/Shh signaling pathway, where it serves / - ^ a s a receptor for SHH/Shh. Inactivating mutations in the PTCHl/Ptchl gene result in a JL pathological activation of the pathway, which may lead to several forms of familial and sporadic cancers. Although PTCHl/Ptchl is generally supposed to be a tumor suppressor gene, mutational inactivation of one allele seems to be sufficient for tumor formation. In this chapter the current knowledge about the mode of tumor formation caused by PTCHl/Ptchl mutations is presented. Most of the insights in these processes are derived from Ptchl mutant mice. The impact of the genetic background on tumor susceptibility, the involvement of additional molecular pathways and epigenetic mechanisms leading to tumor formation, are discussed.

Introduction PTCHl/Ptchl is a key modulator of signaling in the Sonic hedgehog (SHH/Shh) pathway. It is predicted to contain 12 transmembrane-spanning domains and two large extracellular loops.^ Within the plasma membrane PTCHl/Ptchl forms a receptor complex with its signaling partner Smoothened (SMO/Smo), which transduces SHH/Shh signaling (Fig. 1). The signaling pathway controls cell differentiation and proliferation and plays a prominent role in patterning of several organs during embryogenesis. The physiological activation is induced by SHH/Shh. Binding of SHH/Shh to PTCHl/Ptchl suspends the inhibition of SMO/Smo, which leads to signal transduction resulting in induction of target genes such as GUI/GUI and PTCHl/Ptchl itself. Pathologically, the pathway can be switched on by activating mutations in SHH/Shh or SMO/Smo or by mutational inactivation oiPTCHl/Ptchf (Figs. 1 and 2). In humans, inherited mutations in PTCHl result in the nevoid basal cell carcinoma syndrome (NBCCS), which is an autosomal dominant disorder characterized by a combination of developmental defects with a predisposition to tumor formation.^' The clinical findings in NBCCS include a generalized overgrowth of the body, closure defects of the neural tube, and skeletal abnormalities such as bifid or missing ribs and Polydactyly.^ The most frequent tumor in NBCCS is basalioma (BCC), which is usually located in sun-exposed skin areas and may be very pleioformic (Fig. 3). Less frequent, but clinically more problematic are medulloblastoma (MB), which develop in approximately 5% of the patients, and other tumors, including meningeoma, fibroma, and rhabdomyosarcoma (RMS).

*Heidi Hahn—Înstitute of Human Genetics, University of Gottingen, Heinrich-Duker-Weg 12, Gottingen D-37073, Germany. Email: [email protected].

Hedgehog-Gli Signaling in Human Disease, edited by Ariel Ruiz i Altaba. ©2006 Eurekah.com and Springer Science+Business Media.


54

physiologically activated

mactive

SHH:)

^-j?

pathologically activated

Eft

.SH

/

Figure 1. SHH/Shh signaling pathway. Left: in the absence of SHH/Shh, PTCH1 /Ptch 1 blocks the aaivity of SMO/Smo. Middle: during embryogenesis activation of the pathway can be triggered by binding of SHH/Shh to PTCH 1/Ptch 1. Right: activation of the pathway by aaivating mutations in SHH/Shh or SMO/Smo or by loss of ftinction PTCH 1/Ptch 1 mutations (mutations are indicated as stars). Mutations in PTCH 1/Ptch 1 are the most frequent cause of tumorigenesis. The two available strains of mice heterozygous for Ptchf'^ have proved a valuable tool to investigate the role oiPtchl in the control of normal cell proliferation and in neoplastic transformation. In the Ptchl^'° strain/ exons 1 and 2 of Ptchl have been replaced by the LacZ gene, whereas in Ptchr'''^'^'^ mice exon 6 and 7 are missing. As in human tumors associated with PTCHl mutations, mutational inactivation o^Ptchl results in a pathological activation of its signaling pathway with consecutive expression of high levels of GUI and Ptchl mRNA^'^ (Fig. 2). Tumors that arise in these animals are the same as in patients with NBCCS and include MB and RMS (Fig. 4). Although BCC do not develop spontaneously in these mice, they can be induced by irradiation. ^^'^^ Since the initial discovery of inherited PTCHl/Ptchl mutations, the role of this gene in the pathogenesis of a range of sporadic tumors has been the subject of intensive studies. Mutations in PTCHl have been detected in sporadic BCC, which is the most common sporadic tumor in humans with more than 6 0 0 000 cases per year alone in the United States. ^^ BCC are semimalignant and they rarely metastasize, but they can lead to extensive morbidity due to local tissue destruction. The majority of BCC, both sporadic and hereditary, have been linked to the PTCHl locus ^^ and biallelic inactivation oi PTCHl has been detected in a subset of these tumors. ^ Inactivating mutations in PTCHl have also been found in sporadic MB, which is a very malignant primitive neuroectodermal tumor of the cerebellum and accounts for one-quarter of all intracranial tumors in children. ^^' The list of sporadic tumors with mutations in PTCHl today includes also trichoepitheliomas, esophageal carcinoma, bladder cancer and other cancers, although PTCHl mutations account for only a fraction of these tumors.

RMS

€•11 Ptchl wmm

Glil

Gapd

SM

Figure 2. Activating mutations in SHH/Shh or SMO/Smo or mutational inactivation oiPTCHl/Ptchl result in increased levels oi GUI/GUI and PTCHl/Ptchl mRNA in tumor tissue. RT-PCR shown was performed on RNA isolated from murine normal skeletal muscle (SM) and rhabdomyosarcoma (RMS).

Mode o/PTCH 1 /Ptch 1 -Associated Tumor Formation

55

îiititiitiiiiiittitri Figure 3. Basal cell carcinomas of the eye lid. Tumor formation can be also caused by mutations in other members of the SHH signalling pathway including SHH and SMO, or by a deregulation of the signaling cascade due to other mechanisms. Thus activating SMO mutations have been detected in up to 20% of sporadic BCC and recently, SHH mutations have been reported in BCC from patients with Xeroderma pigmentosum.^^ Abnormal expression of SHH seems to be involved in the formation of tumors such as pancreatic cancer, digestive tract tumors and in a subset of small-cell lung cancers. ' The focus of this chapter is on tumorigenesis associated with PTCH 1/Ptch 1 mutations and on the genetic modifiers of this process.

Is Loss of Both PTCHl/Ptchl Formation?

Alleles Required for Tumor

PTCHl is widely assumed to be a tumor suppressor gene. Tumor suppressor genes normally exert a negative control on cell growth and the accepted paradigm is that inactivation of both alleles is required for tumor formation.^ For BCC, it has been proposed that mutations in PTCHl are necessary for passing the genetic "threshold" of the neoplastic process and that PTCHl plays a gatekeeper function in BCC development.^^ Following inactivation of both

56


rhabdomyosarcoma

normal cerebellum

meduUoblastoma

Figure 4. Tumors in Ptchr'^^'^ mice. Left top: macroscopic view of a rhabdomyosarcoma of the back/ hindleg (arrow indicates the tumor); left bottom: microscopic (H&E staining) view of the same tumor. Right top: healthy mouse and its normal brain; right bottom: characteristic bearing of the head of an animal with meduUoblastoma; gross appearance of the tumor. alleles in one cell, clonal expansion occurs, accompanied by the accumulation of further, genetic events, which likely include mutations in genes such as TP53 or RAS}^ However, only 6 0 % of sporadic B C C show L O H at the PTCHl locus,^^ for half of which loss of both PTCHl alleles has been demonstrated.^ T h e same applies to M B , where 12% tumors show L O H at the PTCHl locus but only 7 % have additional inactivating mutations in the coding region o f P r C / / / . 2 ^ - ^ ^ T h u s , not all tumors with L O H at the PTCHl locus show mutations in the remaining wild type allele, although it should be noted that most screens hitherto have covered only the coding region o^ PTCHl y leaving out possible PTCHl mutations in the regulatory elements of the PTCHl gene. Similar observations have been made in tumors found in PtchT^"^'^'^ or Ptchl^'^^^'^ heterozygous mice. As in many h u m a n PTCHl-2issoc\2Lted tumors, the wild type Ptchl allele is retained in M B and RMS of these mice. Also in nodular BCC, which can be induced in PtchP^ mice bv ionizing radiation, retention of the normal Ptchl allele is detected in all tumors examined. Only the infiltrative B C C in these mice show loss of wild type alleles. Similarly to most PTC///-associated tumors in humans, all tumors of Ptchl mutant mice overexpress GUI and Ptchl transcripts. ' O n the other hand, loss of one PTCHl allele is not automatically sufficient for tumorigenic transformation of every cell accompanied bv overexpression of GLIl and PTCHl in patients with N B C C S (i.e., PTCHl heterozygous).^ W h a t is the explanation for this discrepancy? T h e answer may be emerging from recent Ptchl analyses of tumors in Ptchl heterozygous mice. T h e mechanism of overexpression oi GLIl IGUI and PTCHl/Ptchl transcripts in tumors with both PTCHl/Ptchl alleles inactivated is well established. PTCHllPtchl regulates itself via a negative feedback mechanism. Loss of P T C H l / P t c h l function results in a transcriptional activation of GLIl/GUU protein product of which binds to G L I l / G l i l binding sites in the PTCHllPtchl promoter. This results in a transcriptional activation of the PTCHllPtchl gene and an overexpression o^ PTCHllPtchl transcripts^^'^^ (Fig. 2).

Mode ofVl^CWllVtchl-Associated Tumor Formation

57

Prc/i7'^'^ transcript

A

Exon5 Exon 5

Kxon 6

Kxon 7

1 ^

Exon6

[ixon 8 T

C

C

T

^

G

•••-.^^'

G

200

zL

T

t

A A/

Ptchl wild type locus 10

PtchlA^^^ transcript B

Exon 5 T Exon 5

C

C

1 T

C

Exon 8 G

C

T

J

Fxon 8

PtchP'''''^''^ \ocm

AA

A . A . AI

/ \/ \ A / \/\

wt

M7

1

RMS

\/\> \l

Figure 5. (A Wild type and B) mutant Ptchl locus of heterozygous Ptchl"^'^^'^^^ knock-out mice. The neo-cassette of the targeted Pfrî"^ allele is excised and exon 5 is spliced together inframewith exon 8 of the murine Ptchl gene. The resulting PftrAi"^^^transcript is highly overexpressed in RMS o^PtchV^"^^'^ mice as demonstrated in C). In RMS of Ptchl^^^ ^ ^ mice the observed overexpression of Ptchl is restricted to Ptchl -transcripts, i.e to transcripts derived from the allele mutated by homologous recombination. In these transcripts, the neo cassette is spliced out and exon 5 is spliced into exon 8 (Fig. 5). The corresponding data from tumors derived from Ptchl^^ mice, in which exons 1 and 2 have been replaced by a lacZneo cassette,^ are more equivocal. One report showed no expression of wild type Ptchl in cell lines derived from meduUoblastoma of these mice. In contrast, earlier reports about the same tumors showed Northern blot analysis and in situ hybridization which indicated that wild type Ptchl transcripts are expressed in medulloblastoma.^ One way to explain this discrepancy is that Northern and in situ hybridization techniques are semi-quantitative and a repression of Ptchl wild-type transcripts in these tumors in comparison to normal cerebellum may have been overlooked. The existence of similar asymmetry of allelic usage in human tumors has not been investigated. Why is the overexpression oi Ptchl restricted to only one Ptchl allele? One possible explanation is a selective epigenetic block of the wild type Ptchl allele in tumor tissue, a mechanism that has been described in other tumor suppressor genes. Examples include the genes VHL, MLHl OT pl^^ and the predominant underlying mechanism appears to be methylationtriggered transcriptional suppression.^^ Similar mechanism could result in a complete loss or in a repression of the wild type Ptchl allele. In support of a methylation involvement, Herman and colleagues reported that treatment with the DNA demethylating agent 5-azacytidine (5-azaC) increased wild type Ptchl mRNA expression in cell lines derived from medulloblastoma in murine Ptchl^^^ ^ heterozygotes. The location of the responsible regulatory elements in the PTCHllPtchl gene is at present unknown, as is the mechanism restricting the silencing to the wild-type allele. In any case, in consequence of transcriptional repression, the levels of functional Ptchl protein most likely drop below a distinct threshold, finally resulting in tumor formation. These recent observations suggest the therapeutically exciting possibility of restoring PTCHl expression by demethylating agents. In conclusion, mutational inactivation of both PTCHl/Ptch alleles apparendy is not required for tumor formation, although it occurs in a subset of human and murine tumors. In

58


BCC induced by irradiation oiPtchr'^^^' mice loss of both alleles rather seems to be characteristic for later stages of the tumor. Repression of the remaining wild type allele, possibly by methylation, may be involved in the formation of RMS and MB in Ptchl^^^ ^ ^ mice. Both mechanisms—mutational inactivation of both alleles or mutation of one PTCHl/Ptchl allele and repression of the remaining wild type PTCHl/Ptchl allele—are expected to have the same effect, i.e., a deregidation of the SHH/Shh signaling pathway. Alternatively, haploinsufFiciency in PTCHl/Ptchl may be sufficient to trigger tumor formation in combination with mutations in other genes of the SHH/Shh pathway and other pathways, although experimental evidence is lacking. There is, however, increasing knowledge that inter-pathway interactions modify the penetrance and expressivity of tumors associated with PTCHl/Ptchl mutations.

What Are the Modifiers of PTCHl/Ptchl-Associated

Tumorigenesis?

Evidence from Mouse-Strain Comparisons Although inherited inactivating mutations in PTCHl are highly penetrant,^ their expressivity is highly variable. ^ Thus, patients with the same molecular lesion can exhibit very different symptoms, suggesting either a strong impact of environmental factors or the existence of genetic modifiers. The detection of cancer susceptibility modifiers in humans is a difficult and challenging task, ^' and these modifiers remain unknown for PTC///-associated tumors. In contrast, there is emerging evidence for genetic modifiers of tumorigenesis in Ptchl mutant mice. Both the incidence and type of tumors in Ptchl mutants vary considerably with their genetic background. The C57BL/6 (B6) strain is susceptible to MB, which develop in 45% oiPtchr'"^'^'^ mice, but is completely resistant to development of RMS. In contrast, RMS develop in more dian 50% o{Ptchr''^'^N] x B6 or Balb/c x B6 Fl mice. MB is less frequent in these crosses than on the B6 background (21% and 8%, respectively; unpublished data). This implies the existence of either dominant A/J or Balb/c alleles that confer MB resistance and RMS susceptibility, or recessive B6 alleles that confer MB susceptibility and RMS resistance. In addition, initiation as well as progression of BCC in Ptchl mutant mice is influenced by genetic background. Crosses between Ptchl mice and Car-S (FIS) or Car-R mice (FIR), which are susceptible and resistant to skin carcinogenesis, respectively, showed that VlSPtchl mice were highly susceptible to IR-induced BCC, whereas ¥\KPtchl mice were completely resistant. The development of microscopic and macroscopic BCC lesions was also influenced by Car-S and Car-R genotypes, indicating that tumor penetrance as well as initiation and progression of BCC can be modulated by genetic background (Pazzaglia et al. Cancer Research, in press). These results raise the prospect of mapping the genomic loci that are responsible for the tumor susceptibility or tumor resistance. That can be done by QTL (quantitative trait loci) analysis either by intercrossing the F1 mice or by backcrossing them to one of the parental strains. The mapping of the loci is achieved by typing of strain-specific microsatellite markers, which are available for the strains mentioned above. Identification of these modifying genes is relevant to humans, in whom BCC is the commonest malignancy, and MB and RMS are the most common brain tumor and soft tissue sarcoma in children, respectively. Especially for sporadic BCC, a genetic predisposition has been demonstrated in the general population and efforts to identify the susceptibility loci are in progress. ' As mentioned above, the identification of tumor-modifying loci in humans is difficidt. The hope is that QTL studies in Ptchl mutant mice will help to identify genes modifying PTCHl/Ptchl-2LSSoci2ited tumor formation.

Mode o/'PTCH 1 /Ptch 1 -Associated Tumor Formation

Evidence from Gene Expression and from Biochemical Studies Besides QTL, other approaches can be used to identify interacting signaling molecules that might contribute to the malignant transformation of a heterozygous PTCH 1/Ptch 1 +/- cell. Several labs have performed gene expression analyses oi PTCH 1/Ptch 1 associated tumors or of cells overexpressing GLIl/Glil or treated with SHH/Shh. It has been shown that the expression of the proto-oncogene Nmyc can be stimulated by Shh in cerebellar granule neuron precursors and that Nmyc is required for the full Shh proliferative response of these cells. Nmyc is involved in cell cycle progression and is implicated in repression of genes that promote cell cycle exit. Interestingly, Nmyc is expressed at high levels in MB. It also has been demonstrated that the expression oiPDGFRa is activated by GUl.^^The expression of PDGFRa is accompanied by the activation of the Ras/Raf/Mapk pathway, which regulates cell proliferation. The induction of PDGFRa by GLIl/Glil is likely to occur in vivo, since PDGFRa is overexpressed in human and murine BCC. '^ Furthermore, PTCHl binds to, and sequesters Cyclin B1 and thus inhibits cell cycle progression. It has been speculated that some PTCHl/Ptchl mutations increases cell proliferation by releasing Cyclin B. ^ Shh also promotes the transcription of Cyclin E and Cyclin D, two inhibitors of Rb and principal regulators of the cell cycle. ^'^^ The up regulation of Cyclin E expression was accomplished through binding of Glil to the Cyclin E promoteT, which mediated the ability of Shh to induce DNA replication. Upregulation of cyclin D expression by Shh may be responsible for the distinct ability of Shh to promote cellular growth. Altogether, it is likely that tumor formation is contributed to by a proliferative advantage resulting from the induction of Nmyc, Cyclin B, Cyclin D, Cyclin E and/or PDGFRa expression. Gene expression analyses also revealed an interaction between the tumor-promoting IgfZ and Shh signaling. Tumors in Ptchl heterozygotes consistendy overexpress Igf2. In addition, Ptchl mice deficient for Igf2 did not develop MB and RMS, which demonstrates that Igp2 is indispensable for tumor formation. Since the overexpression oÎgft in tumors oiPtchl mice seemed to be independent from genetic or epigenetic changes, it has been proposed that Igf2 is a direct target of Shh signaling pathway. The following observations suggest that an important function of Igf2 in RMS formation in mice is to activate the PI3K/Akt/PKB signaling pathway (Fig. 6): Igf2 usually activates receptor tyrosine kinases such as the Igf-1 receptor or the insulin receptor. Activation of these receptors both increases mitogenesis and inhibits apoptosis, which is attributed primarily to the activation of the Ras/Raf/Mapk pathway and of the PI3K/Akt/PKB kinase pathway, respectively.^^ However, the relative importance and contribution of each pathway to tumor growth is unknown. The phosphorylation status of the growth-stimulating Mapk pathway was unchanged in tumors in Ptchl heterozygous mice, whereas the cell-survival-promoting PI3K/Akt/PKB signaling pathway was highly activated.^ Furthermore the cell cycle inhibitor p27kipi ^j^j ^j^g growth inhibitor Gadd45a were up regulated in these tumors. In agreement with this result, a low proliferation rate of RMS cells was demonstrated using the cell proliferation marker Ki67. In addition, the anti-apoptotic Bcl-2 protein was overexpressed in tumor cells whereas the expression of the pro-apoptotic Bax was low. These results show that besides induction of proliferative signals (see above) deregulation of the SHH/Shh pathway also seems to induce anti-apoptotic signals. In agreement with these observations, Bcl-2 and Gadd45(X have been recently shown to be direct targets of Glil^^(Kappler et al. Int. J. of Oncology, in press). A model is conceivable in which elevated levels of Glil increase the expression of the anti-apoptotic Bcl-2, of the growth inhibitor Gadd45a and of the tumor-promoting factor IgfZ. That, together with the activation of the anti-apoptotic PI3-kinase signaling pathway, may decrease apoptosis and enhance tumor growth (Fig. 6).

59

60


pteii

proliferation^

U0

sâpoptosis

Figure 6. Delineation of the molecular events in RMS oiPtchr''^^^^ mice. Mutation in Ptchl is indicated by a star. Our data indicate that anti-apoptotic signals predominate and that proliferative signals in these tumors are only weak. PTCHl/Ptchl also induces apoptotic cell death via increased caspase activity unless its ligand SHH/Shh is present to block the signal. 5^ Therefore P T C H l / P t c h l may be a pro-apoptotic dependence receptor. Dependence receptors transduce two different intracellular signals: in the presence of a ligand, they transduce a positive signal leading to cell survival, differentiation or migration. In the absence of a ligand they initiate or amplify signals for programmed cell death. ^^ The finding that PTCHl/Ptchl is a dependence receptor may have important implications in the understanding of P7T7/////V

T3

5 Si i5 6 E «/5 0)

O

03 ^

a;

2

oj

Q- i :

Q.

c (D fl)

O c

a. o Z

O

z

t s

> (^

t:

^ -^^ ^ ^

C

'-^ —I

C ^'03

3 ^H

^ 5^ 5 9^ 5 ^

I A

9^ o cy^
= X

T£

^

03

.

'^ . i£ i^ U CD

ait; Q.-D CD CD

OJ

s

Q.

t« M—

"D m CD CD 0) "D tr> DO 03 O

CD CD

E u u b js >L X E (U D D O)

^

53 Ô

Q_

o C C "O JO t/5 O

ns

S £

•^

JC

U

CD

^-^

DO i_

O

•o ^ DO •3 C -O

C

—^

^

C Jf ^

0

Q-

a;*s _o _^> c c O ***) .9.—J2 _^ _i^ ^

03

on

c O Q . Q-

Q- O

O O

5 5

Q

It IIII

1 . 1.

o Q_ 1-

O

3

•s

I

<S f5 r^ f5

E rn

.':: rD .\^ r^

5Q E o

5 S2

CD

N

3

3

3

o

o

o

5

2

1. 55

3

91

splitting Hairs

7^ -O

s "s t -. 4£

G^ T-

C^ O^

CO LD

IT)

-D

3

n3

. X ^ X - a: X ^ cil X ^ . T^" ."i h N o J U o U o - ' i ^ C J

X m o^' m o rv oS' ,^r

vX5

Q_

CD

•fc

o

O

E

Q-

CD

1-s

O

a;

E o n

CD •a

0)

^ _>>

a; u *n

Q-T:

> (\)

"O

O C

CD

OS t/5

b O

o

03

t/5 CD

O

Q.

C

O

Q_ t/5 9-

•
s -c ^ ^ -c^

CD DJO OS

CD

'^. -c -c (D Q_

^ U OS

in

^ Z

z zz u

c:^ uj

P

CD

>— GO

.—

§-^

o_ o =

o

T-

^2

U

li

^ -^

CD 0) 3 3

o

Wl CD *" CD

^ t5 t ^11

o r^ o o.^ o.^

E *E

E? O

m

^

^ cj^

OS

>>

CD

o ^

Q . CD • ^ U C Q; E -S^ ^ O ,CD Z CO I—

OS

3 C/5

c

CD

~B °-

CD CD CD _D JD - D 3 D D

3

O §

^

Q^ Z

CD

CD

CD

Q.

>

ti

(D

o X

CD

t/5

1^

Q_

o _a;

0)

J ^ S^ —J

CD CD

— «/> Jf >>

o o o

> -F

t3

O

CD T 5 Q _ CD

M— ^ + - M_

O E

DJO C

CD

"g.i 2c E Q5

^ g:

o

"D

(D

2 ^

CD

QJ

n c o cu -a ,o "TJ (1)

Q_

Q

O

OS

U

s .2 i_

^ .r: CD U

01

? o

C-

0)

D

2 <M

CM

J

o E

z
o>

u

UJ

z

E

CU t/3

O

E

I

rn

< u u CN

o o

CD

CD

13

3

o o E E

U 13

ti u

ti u

1 1 1 L U

I I I U J

rn c

' "

GO O

c

03

E CU

u

D

E

n

trt

>

M

î2 M2

'^ _o cOi > > (- > o ci;

Q.

CN

c

0)

GO QJ

O

o

O

L. 03 Q_ X CD

o

' it

T3 O _C ^ QD T3

u ^

o v2

.9 2 2

O •a Q. O c O (T! ^

u

M~

CD

Q_ 03 X r-

O

_D T5

.y Q.

Q_

U 13

O • ^

3

94


Shh /

1 \

/ V/ \

x--x^^

/îi^&(21^

^^£\^

Class 1

Class II

Class III

/.e. ?

/.e. Ptci

/.a M l

CydiInD

Wmmm

Figure 1. Model ofGli-mediated transcriptional control of Hh target genes. Despite common Gli consensus sequence, three classes of Hh target genes maybe differentially regulated by Gli functions. The level of Hh signaling received by a target cell will determine the balance of Gli activator and repressor functions."^^ The balance of Gli functions will influence expression of target genes differendy, where specificity of the Gli consensus site may be modulated by other aV-elements or transcriptional cofactors. Expression of Class I genes will be controlled solely by Gli3 repressor function. Class II promoters will be influenced bv both Gli aaivator and repressor functions; basal expression can be derepressed in low Hh signals (no G l i ^ , while high level induction requires high Hh-dependent Gli flinaion. Class III genes will only be induced by Gli activator functions. what modifications are required to generate Gli2^^, as well as whether Gli/Ci activators and repressors regulate expression of the same set of target genes.^^'^^ While both activator and repressor forms bind the same consensus sequence in vitro and in vivo, it appears that the sensitivity of particular target promoters to regulation bv either activator or repressor functions is determined by m-regulatory elements (Fig. l).^^'^^'^

Gli Genes Have Unique and Overlapping Functions in Vivo Genetic studies suggest that Gli2 and Gli3 are the primary mediators of H h responses in mammals, where Gli2 functions as the principal activator and Gli3 performs most repressor functions during normal development. Despite its potent activator function and its induction by H h signaling, Glil is likely not involved in initial H h responses. GUI mice show no apparent developmental defects or alterations in H h responses, unless one copy of Gli2 is removed. In contrast, mice die in late gestation due to severe defects in spinal cord development, as well as developmental abnormalities of lung, vertebrae and bones, with global down-regulation of H h target gene expression. ^^' The importance of Gli2-dependent activator function is further revealed in PtcV ; Gli2 double mutants, where Ptcl posterior neural tube defects due to hyperactive H h signaling are rescued by removing Gli2.^^The importance of Gli2 repressor function during normal development is unclear, as Glil can fiilly substitute for Gli2 activities embryonically. However, G//^ adults do develop progressive skin defects consistent with deregulated H h responses. In contrast, loss of Gli3 repressor results in ectopic activation of H h responses contributing to profound skeletal and nervous system defects found in Gli3 embryos, which die perinatally. ^^^ Controlling Gli3 repressor activities is a key function of Shh signaling. Genetic reduction of Gli3 repressor function in

Spliuing Hairs

95

ShhGlid compound mutants displays a dose-dependent rescue of the severe growth arrest, ventral spinal cord patterning defects and distal limb defects observed in Shh mutants. '^^'^ While Gli proteins have distinct activities, their overlapping functions are revealed in compound mutants. Defects in neuronal specification in GUI Gli2 spinal cords are more severe than in the 67/2"^'counterparts, ' indicating that GUI can functionally substitute for Gli2 in the spinal cord. One dose of Gli3 is sufFicient to induce target genes in Gli2 \ Gli3 presomitic mesoderm,^^' indicating Gli3 can substitute for Gli2 activator function in vivo. Furdiermore, Gli:f^^'^^^ mutant mice show some signs of rescue in the neural tube, suggesting that Gli3 can function in vivo as an activator, albeit less potendy. Due to their dual role as transcriptional activators and repressors, loss of Gli gene functions is not functionally equivalent to loss of Hh signaling. Defects observed in somites, neural tube, lung and trachea of 67/2" mutants likely reflect the simultaneous loss of positive Hh responses and inhibitory repressor flinctions.53'58.59,66,67

Divergence within ShhIGli Signaling Cascade Whether members of the Gli/Ci transcription family are the sole mediators of Hh responses in target cells has been a controversial issue. In the larval fly eye, Hh signaling through Ptc and Smo is required for development, although target gene transcription occurs independently of Ci function. Similarly, the minimal Hh-dependent regulatory regions of the wg promoter in flies and the COUPTFIIipTovaottr:in mice do not contain consensus Gli binding sites. '^^ However, recent work in both flies and mammals support that most, if not all, of Hh responses are Gli/Ci-dependent.^'^'^^'^^'^^ In the ventral spinal cord, patterning of all neuronal subtypes by the ventral-to-dorsal gradient of Shh is dependent solely on Gli function.^^'^^' These studies suggest that most cases of apparent Gli/Ci-independent Hh responses can be attributed to derepression of Hh target gene expression in the absence of both Gli/Ci activator and repressor functions. Another possible branch of the Hh/Gli signaling pathway comes from studies suggesting that there are some Gli functions independent of Hh signaling. Gli2 and Gli3 are expressed more broadly than sources of Hh signals during development.^^ Although Gli repressor functions may be required in these regions to inhibit inappropriate activation of Hh targets, it is also possible that the activities of Gli proteins are subject to other signaling inputs. During Xenopus development, FGF activates Gli2 and Gli3 expression, which are required to maintain and pattern mesoderm through induction of key regulatory genes, including Brachyury antagonistic relationship between Hh and BMPs (Bone Morphogenetic Proteins) observed during development may involve direct interaction of truncated Gli3 repressor and Smads, the mediators of BMP signaling. While recent studies have revealed cooperative functions for the Gli transcription factors in several developmental contexts, their distinct and overlapping roles in mediating Hh responses in the skin was not known. Given the importance of regulated Hh signaling during skin development and tumorigenesis,^'^'^ understanding the functional interplay of Gli transcription factors and the transcriptional cascades they control in the skin is necessary.

Molecular Mechanisms of Embryonic and Adult Hair Development Epidermal-Dermal Signaling during Embryonic Hair Follicle Morphogenesis Despite their differences, epidermal appendages, including mammalian hair follicles or avian feathers, share a similar developmental programme. A reciprocal exchange of inductive cues occurs between mid-gestation ectoderm and mesenchyme, to determine where these structures will form and trigger their initial development (Fig. 2). Recombination experiments suggested the first signal arising from the mesenchyme instructs regions of the pluripotent ectoderm to elongate into regularly spaced placode structures.^ Epidermal signals from the placode cue underlying mesenchymal fibroblasts to organize into dermal condensates, or


96

ectodtemn

mesenchyme

Embryonic Hair Foilicle Morphogenesis

v - ' j i ^ matrix

s dermal si^f^l s epldemiai signal

Catagen

\\\ " * \ v " basaloid . / / ' *' tumor nest

Figure 2. Initiation of embryonic and adult anagen hair follicle development proceeds through inductive epidermal-mesenchymal signals: Embryonic dermis instructs overlying ectoderm to initiate placode formation, a structure which signals to mesenchymal fibroblasts to form dermal condensate. Subsequendy, the dermal condensate signals to follicular keratinocytes to stimidate proliferation and downgrowth into developing dermis. The dermal condensate will be enrobed by epithelial cells and mature to form dermal papilla. Proliferating hair progenitors (matrix cells) that are pushed away from contacting the dermal papilla will differentiate in response to lateral signaling between keratinocytes into the 6 epithelial cell types making up the hair shaft. Shh is expressed in early placode and remains expressed at the distal tip of developing follicles. Adult hair follicles undergo cyclic periods of growth (anagen), destruction (catagen) and rest (telogen). Matrix cells are thought to maintain their proliferative capacity through close association with dermal papilla during anagen. During catagen, withdrawal of growth signals results in massive apoptosis of the lower two thirds of the epidermal hair follicle. As a result, the dermal papilla is dragged through the dermis and comes to rest close to the permanent bulge region (telogen), during which the old hair shaft may be shed. At the onset of a new anagen phase, epidermal stem cells are stimulated to divide in response to signals from the dermal papilla. Epidermal progenitors proliferate to regenerate the matrix cells and their differentiated progeny necessary for a functional hair. Transient Shh signaling is required during the telogen-anagen transition and deregulated Hh responses can result in uncontrolled progenitor growth in BCC tumors. Key: ORS: outer root sheath; IRS: inner root sheath; IFE: interfoUicular epidermis.

splitting Hairs

97

dermal prepapilla. A second "dermal message" from the condensate subsequently prompts epidermal cells in the placode to proliferate and grow down into the dermis, forming a hair peg, which will eventually engulf the dermal condensate to form a dermal papilla. This close epidermal-mesenchymal crosstalk is believed to stimulate further proliferation and differentiation of epidermal cells into many components of the mature hair foUicle.^^ In the hair peg, epidermal cells that lose contact with the dermal papilla become outer root sheath (ORS) cells, a compartment that is contiguous with the basal layer of the interfoUicular epidermis (IFE).^^ Matrix cells are highly proliferative epidermal cells that maintain contact with the dermal papilla and start to adopt distinct differentiation programs on withdrawal from the cell cycle. These differentiating cells move upwards in morphologically distinct concentric cylinders of cells that emerge at staggered intervals during postnatal development; the outer three cylinders of the inner root sheath (IRS) develop first, while the three central hair shaft layers arise later. The final epidermal lineage to emerge is the oil-rich sebocytes, which populate the sebaceous glands, off the upper portion of hair follicles.

Adult Hair Cycle Many aspects of embryonic epithelial-mesenchymal crosstalk are redeployed in adult skin where hair follicles cycle through periods of active grovnh (anagen), periods of regression (catagen) and rest (telogen) (Fig. 2). During anagen, matrix cells enrobing the dermal papilla continue to proliferate and differentiate, elongating the emerging hair shaft. Matrix cells have a finite proliferative capacity, influenced in part by the size and stimulatory output of the dermal papilla.^ '^ When the proliferative potential of matrix cells is exhausted, a destructive phase ensues resulting in apoptotic cell death of the lower two-thirds of the hair foUicle.^"^ During catagen, the dermal papilla also dramatically reduces in size and transforms into a quiescent cluster of cells. However, the shrunken dermal papilla is pulled through the dermis trailing the lower portion of the follicle as it regresses to a resting position close to the permanent bulge region, the niche of epidermal stem cells.^'^^ The resting telogen hair follicle remains essentially inactive, with a small ORS-like hair germ and rudimentary dermal papilla, separated by a thick basement membrane.^^ Having lost its anchorage, the old hair shaft:, or hair club, may be readily shed.^'^ Although the exact nature of the trigger to enter the next anagen cycle remains unknown, it is hypothesized that in response to signals from dermal papilla, epidermal stem cells in the bulge region are transiendy activated to proliferate and repopulate the lower portion of the hair follicle ("bulge activation hypothesis"). '^^ Subsequendy, dermal progenitors are recruited to the dermal papilla from reservoirs in the adjacent dermal sheath^ likely in response to epidermal signals from secondary hair germ. As anagen proceeds, a similar cascade of mesenchymal-epithelial cross-talk that leads to the production of matrix cells, inner root sheath and hair shaft lineages during embryonic hair follicle morphogenesis is redeployed.

Shh Signaling in Hair Follicle Development A putative candidate for the first epidermal signal required for induction of the dermal condensate is Shh, expressed early in the developing placodes.^'^'^^ Although overexpression of Shh can induce ectopic placode formation in chick skin^ '^^ Shh overexpression in embryonic mouse skin results in disorganized hyperproliferative epithelial growths.^ Furthermore, hair follicle induction is initiated in Shh mutants, but subsequendy arrests, resulting in small placodes with rudimentary mesenchymal condensates.^'^' These findings suggest that Shh signaling is not required for initial specification events, but plays an important role in subsequent signaling required for dermal papilla development^^ and/or hair follicle downgrowth.^ Consistent with this, expression of Hh target genes GUI and Ptcl is detected in both epidermal and mesenchymal compartments of hair follicles throughout development.^''^ Rudimentary Shh'' dermal papilla contain fewer PDGFR-a-positive mesenchymal cells and fail to express differentiation markers, such as alkaline phosphatase and Wnt5cL^ suggesting that Shh signaling is required for proliferation and differentiation of papilla fibroblasts. '^^'^^ Epithelial-epithelial Shh signaling is required to promote proliferation of epidermal cells in developing hair

98


follicles. ' Furthermore, terminal differentiation of interfoUicular keratinocytes is unaffected in Shh mutants, while disorganized hair follicle structures express low levels of hair keratins, indicating that Shh signaling is likely not required for epidermal lineage determination, but may regulate expansion of hair follicle progenitors. ' ' During the adult hair cycle, Shh is induced in secondary hair germs at the telogen-anagen transition where transient activation of Ptcl is observed in the bulge region. As anagen proceeds, asymmetric Shh expression in distal matrix cells is believed to promote proliferation of hair progenitor cells. Shh induces precocious anagen from telogen follicles'^ while blocking Shh signaling results in telogen arrest of postnatal hair follicles with reversible hair loss. Interestingly, a novel role for Hh signaling in the maturation of postnatal sebocytes has recendy been reported and is likely mediated through Ihh. ' Shh signaling plays an important role in expanding epidermal progenitors and promoting dermal papilla maturation, likely to enable other signaling pathways to specify commitment to hair follicle lineages during embryonic and adult hair follicle development.

A Fine Balance: Gli Activator and Repressor Functions in Skin Development During mouse embryonic hair follicle development, Gli2 is the major activator mediating Shh responses in the developing skin.^^ The Gli2 follicle defect phenocopies the arrested follicular development of 5 M " mutant mice, with reduced Shh target responsive gene expression and decreased cell proliferation. Through tissue-specific transgenic rescue experiments in mutant background, we have shown that activation of Gli2 function in the epidermis by Shh signaling is essential for embryonic hair follicle development. Although the development of dermal papillae is severely affected in Gli2'' skin and Gli2'" dermal fibroblasts fail to respond to Shh in vitro, restoration of Gli2 activity in the epithelium is sufficient to fully rescue hair follicle development in GliZ^' skin. Despite phenotypic similarities, epidermal overexpression of Gli2 does not rescue 5M-/-follicles, suggesting that Gli2 activator function is Shh-dependent. ANGli2, a mutant form of Gli2 lacking the N-terminal repression domain, activates Shh target gene expression constitutively both in vitro and in vivo."^^ Importantly, ANGli2, but not Gli2, can induce GUI and Ptcl expression in Shh'' skin and induce ectopic expression of GUI and Ptcl in regions normally lacking Shh response (Fig. 3). These results establish that ANGli2 functions as a constitutive activator of Hh target genes in vivo. Similarly, in Drosophila, the activator function of Ci has been shown to be Hh-dependent; an uncleavable form of Ci, which is incapable of forming the repressor form, cannot induce the expression of target genes in the absence of Hh signaling. While these results demonstrate the Shh-dependent activator function of Gli2, it remains unclear whether Gli2 also acts as a repressor in the absence of Shh signaling in the skin. Interestingly, Glil can compensate for the lack of Gli2 function in GU2 ^ mice during embryonic development, but remarkably, adult GU2p^'^'^^^ mice develop skin defects, including alopecia and ulcers. The molecular mechanisms behind this adult skin phenotype await further investigation, it is plausible that Gli2 may possess a repressor function, which is lacking in Glil, and this Gli2 repressor function plays a key regulatory role in adult hair cycle. Although ANGli2 can activate target gene expression and restore cell proliferation in Shh' ' epithelium, the rescue of hair follicle development is not complete in K5Cre; Z/APANGU2; Shh''' skin. Restored activator function of Gli2 in K5Cre; Z/APANGU2; Shh''' epithelium may compete with high levels of endogenous Gli3 repressor function for target gene control, hindering hair follicle development. GU2 and Shh mutants, which are both deficient for Gli activator function, display a similar but distinct arrest of hair follicle development; the more severe defects in Shh mutants could be compounded by the additional excess of Gli repressor activity. Consistent with the notion that Shh signaling acts to inhibit the repressor activity of Gli3, there is a dramatic rescue of Shh mutant defects in the developing neural tube, face and forebrain,^ as well as limbs, ' when Gli3 repressor function is removed in Shh; GU3 mutants.

splitting Hairs

99

*^!$

Ir*

Figure 3. The activator function of Gli2 is required for hair foUicle development and is Shh-dependent. HematoxyUn-eosin staining of dorsal skin sections from El 8.5 wild type (a), Gli2^' (b), Shh''' (c), K5Cre; Z/APGli2;Gli2'\m),K5Cre;Z/APGli2;Shh-^-{n),K5Cre;Z/APANGli2;GliZ^-{\i),2^dK5^ Shh'' (v) mice. In situ hybridization analysis of Shh target gene expression, Ptcl (d, g, j , o, r, w, and z) and GUI (e, h, k, p, s, X, and aa). Staining of BrdU (green) and K14 (red) in E18.5 wild type (f), GliZ'' (i), Shh'^ - (1), K5Cre;ZlAPGlil; Glil'' (q), K5Cre;Z/APGli2; Shh''- (t), K5Cre;Z/APANGli2; Glil'- (y), andK5Cre; Z/APANGli2; Shh-'- (bb) skin. Scale bars: 50|lm. Reprinted widi permission from: Mill P, Mo R, Fu H et al. Genes Dev 2003; 17(2):282-294. ©2003 CSHL Press Co. To address the role of Gli repressor activity in the developing skin, we examined the development of vibrissae and pelage follicles in Shh; Gli3 c o m p o u n d mutants. As seen in other systems, excess Gli3 repressor contributes to the Shh m u t a n t skin phenotype, since removal of Gli3 significantly alleviates the developmental arrest o^ Shh mutant follicles (Fig. 4, P.M. and C.C.H., manuscript in preparation). Transcription of early cell cycle regulators, including A/^?w)/r and cyclin D2, is derepressed in Shh'''-, Gli3 mutants, contributing to the partial rescue of keratinocyte proliferation (P.M. and C.c.H., manuscript in preparation). Fiowever, transcription of late Gi regulators is also Shh-dependent requiring additional Gli activator functions to allow^ for robust cell cycle progression and rescue of epithelial proliferative responses during hair follicle development. O u r results suggest that Shh controls proliferation of the embryonic epidermis through a complex transcriptional cascade of multiple cell cycle regulators via both Gli-dependent gene activation and de-repression. There is growing support for a model where the balance of Gli activator and repressor functions d e t e r m i n e Shh responses (i.e., the transcriptional read-out of Shh target genes). It has been shown that overexpression of a constitutive Gli3 repressor can override endogenous Shh signaling and associated activator


100

+/+

Shh-I-

ad (

^

^

K^^^^^^^^^^

Shh'

+/+

|g^___^^^^J

am

"-

1

|j

,

1

• T-SfeiSJi * 3 * ^

1

. 1m

Ptd

\

1

r

Figure 4. Arrest of vibrissae follicle development in Shh mutants is due to excess Gli3 repressor activity, (a-e) A7/whole-mount in situ hybridization of E 14.5 embryonic vibrissae follicles (red arrowheads), in wild type (a), Shh'^' (b), Shh'^'; Gli3^^' (c), Shh'''\ Gli3'^' (d) and Gli3''' (E), with ectopic tracts (blue arrowheads). Sections through El.45 whisker pad of wild type (f,i,l), Shh'^' (g,j,m), and Shh'^'\ 67/3" (h,k,n) embryos. Expression o( Ptcl (1-n) is derepressed in Shh'^'; Gli3'^' vibrissae follicles, while expression of GUI (I-k) is not affected. function in dorsoventral patterning of the spinal cord.^^'^^ Similarly, overexpression of a d o m i n a n t repressor form of Gli2 (ACGli2) is sufficient to override endogenous Shh signaling and block Gli activator-directed hair follicle development.

Dissecting Requirements of Signals and Deciphering Cross-Talk T h e importance of interplay between the key developmental signaling pathwâys, such as Wnt, Notch, BMP and Shh, during skin development and homeostasis merits further investigation. Loss o^ Notch 1 in adult epidermis leads to upregulated Gli2 expression and stabilization of nuclear (J-catenin. A yet unidentified W n t signal is required to induce the placode expression oi Shhy '^^^ v^hich in turn induces Wnt5a expression in dermal papilla^ during development and sustain high level Wnt5a expression in the stroma ofBCCs.^^^ W n t signals can antagonize H h responses through induction o^growth arrest specific gene 1 (gasl) expression, a glycosylphosphaditylinositol-linked membrane glycoprotein v^^hich binds Shh.^^^ Gasl expression is induced surrounding developing epidermal appendages, such as teeth and vibrissae, in a domain distinct from the Shh source and partially overlapping wîth PtcL Although the functional dependence on Wnt signaling in the developing mandible has yet to be demonstrated, Gasl expression may be key to sharpening H h gradients allowîng H h responses to be periodically restricted developing teeth. ^ ^^ These signaling pathwâys individually play important roles in skin development and disease, but just how entwined the complex web of synergistic and antagonistic interactions between Shh and other key signaling cascades really has only begun to be understood.

splitting Hairs

101

Shh Signaling Plays a Conserved Role in the Development of Other Epidermal Appendages The initial inductive steps in the morphogenesis of many epidermal appendages involves similar reciprocal epithelial-mesenchymal signals.^^^ Vibrissae are large tactile follicles, which develop earlier than pelage (dorsal) follicles, in regularly patterned tracts in the mystacial pad. While the extent of vibrissae development in Shh'" mutants is difficult to assess due to craniofacial defects, ^'^'^^ cyclopamine treatment of wild type vibrissae cultures arrests their development and ectopic expression o£Shh can induce supranumary vibrissae follicles,^^^ suggesting a role for Shh signaling in vibrissae follicle development. A proper balance of Gli activity is required to establish or maintain vibrissae tracts. G//2'^'and G/zJ" mutant vibrissae develop correctly (in contrast to Gli2 mutant pellage follicle developmental arrest), albeit in reduced or expanded numbers, respectively. Furthermore, a dosage dependent rescue of the vibrissae patterning defects in Shh mutants is revealed by reducing Gli3 repressor function in Shh'^'; Gli3^' and Shh''; GliJ^" mutants, although the development and downgrowth of these structures is only partially restored (Fig. 4, P.M. and C.c.H., manuscript in preparation). Similar to the arrest of hair follicle development observed in Shh'' mutants, both salivary gland and tooth development are also arrested at early rudiment stages.^^ '^^^ Abnormalities in tooth development primarily arise from compromised planar (epithelial-epithelial) Shh responses, resulting in decreased epidermal proliferation and survival while differentiation proceeds unaffected. '^'^^^ Interestingly, while rudimentary teeth are observed in tissue-specific deletion mutants of Shh and Smoy a more severe phenotype is observed in GliZ ; Gli3''' which lack all signs of molar induction, ^ ' ^ '^' ^ ^ ^ suggesting that both Gli activator and repressor fimctions may be involved in initial establishment and/or maintenance of these structures. In contrast, the development of some other epidermal appendages is less dependent on Shh signaling, as is the case with mammary glands and nails, likely due to redundant expression oi Shh and Ihh such that target gene expression in these Shh" organs is unchanged.^ However, a regulatory role for Hh signaling in postnatal mammary gland during pregnancy and lactation may still exist, as ductal dysplasias and hyperplasias spontaneously arise in G//2" and PtcV mammary glands in response to different hormonal conditions. ^ '^^^

Shh: Stem Cells, Cell Cycle and Cancer Regulation ofHh Responses Is Requiredfor Maintenance of Adult Tissues Deregulated Hh signaling has been reported in various human cancers of the brain (medulloblastomas, glioblastomas), muscle (rhabdomyosarcomas) and skin (BCCs), where Hh signaling may be normally required for proliferation and maintenance of specific adult stem cells. Recently, elevated Hh responses have also been implicated in tumors arising from the upper gastrointestinal tract, including lung and pancreas, as well as from the prostate epithelium, where Hh signals likely play similar regenerative functions. ^^^'^^^ In response to physiological and stress stimuli, Shh signaling is likely required for progenitor cell expansion and tissue renewal, such as in the skin. However, tumors can be induced when these processes are not tighdy regulated. Importantly, sustained Hh signaling is essential for tumor maintenance as specific inhibition of H h responses with cyclopamine is sufficient to block tumor growth.i2«-iî'i^^-i^«

Epidermal Stem Cells and Cancer An important characteristic of postnatal skin and its appendages is that they are constantly turning over; keratinocytes are continually challenged by damaging environmental factors (UV irradiation, pathogens and chemical carcinogens) and physical stresses, which require ongoing repair. By virtue of its self-renewal capacity, epidermal stem cells can divide asymmetrically to sustain the pool of self-renewing cells and generate daughter cells that can terminally

102


differentiate to replenish lost keratinoqês.^ '^^^'^ Physically sheltered in the bulge region of the hair follicle, multipotent stem cells within this niche can repopulate all three epidermal progenitor lineages. '^ ^ As long-term residents of the epidermis, epidermal stem cells are the most likely targets of skin carcinogenesis because they accumulate multiple genetic mutations that may be required for tumor formation.^ ^ However, in some cases, disturbing epidermal homeostasis by disrupting the stem cell compartment may also result in tumorigenesis. A tight regulation of when epidermal progenitors are competent to respond to periodic Shh signals ensures that proliferative responses are regionalized and restricted to hair follicle morphogenesis and anagen phase of the adult hair cycle (Fig. 6).^^^'^ ^' In contrast, deregulated activation of Hh responses in mouse skin is sufficient to induce hair follicle-derived tumors that resemble highly proliferative and undifferentiated masses of epidermal progenitor cells without secondary mutations. '^ ^' Basal cell carcinomas (BCCs) are the most common human malignancy and its incidence continues to increase.^ Inactivating mutations in the \\\3s^2sv patched (PTCHl) gene were initially identified as the cause of nevoid basal cell carcinoma syndrome (NBCCS) which predispose affected individuals to the development of midtiple BCCs.^' Mutations in PTCHl have since been linked to sporadic human BCCs and other hair follicle derived neoplasias. ' These studies provided the first molecular clues to the pathogenesis of these hair follicle derived tumors. Overexpression of Shh throughout the embryonic murine epidermis results in early BCC-like lesions and disturbances in hair follicle development, ' ' emphasizing the importance of restricted Shh signal reception. A loss of temporal and spatial regulation of Shh signaling can lead to the formation of multiple tumors types, including BCCs.^ In humans and mice, activated Hh responses in the skin as a result of loss-of-function Ptcl mutations or gain-of-fimction Smo mutations result in hair follicle-derived tumors.^ '^ The downstream Gli transcriptional mediators of Hh signals likely play a key role in the development of skin tumors (Table 3). A hallmark of BCCs is high level expression of Hh target genes GUI and PtcU in the absence oiShh signal (Fig. 5).^'^^^ Epidermal overexpression of human GUI in transgenic mice alone is sufficient to promote spontaneous tumor formation which resembles human BCC and other hair follicle neoplasias, including cylindromas and trichoblastomas. Transgenic overexpression of GU2 in the epidermis results in spontaneous formation of BCCs, ^"^^ while up regulation of endogenous GU2 in the skin, such as is observed in epidermal-specific NotchV''mioe,^ also induces BCC formation.^^^ Unlike GUI, which lacks the N-terminal repression domain and functions as a constitutive transcriptional activator, Gli2 requires Shh-dependent activation to promote Hh target gene expression and responses. '^^' Given the differences in tumor phenotypes between K5GU1 and K5GU2 mice, we hypothesized that the oncogenic potential of Gli2 requires a Shh-dependent activation trigger. As the expression of Shh in the adult skin is limited to hair follicles during the telogen-anagen transition, ^^^ conversion to Gli2'^^^ in K5GU2 keratinocytes would be spatially and temporally restricted, such that hyperproliferation and transformation events yielding BCC progenitors would occur less frequently. In contrast, A!56^//7-overexpressing keratinocytes are ligand-independent, in that their activity is no longer limited to where and when Shh is expressed during the hair cycle, and could therefore target wider tumor progenitor populations. To determine whether the difference in oncogenic potentials between Glil and Gli2 lay in their Shh-dependent activity, we have examined the effect of Gli2 and ANGli2 overexpression on epidermal homeostasis and tumorigenesis using 3 transgenic mouse models; K5GU2 which develop BCCs ^ and the previously uncharacterized adult phenotypes of K5Cre; Z/APGU2 and K5Cre; Z/APANGU2 mice.^^ The Keratin5 {K5) promoter drives transgene expression in the basal layer of the skin and hair follicles, targeting compartments believed to contain epidermal stem cells. Observations from these studies support three emerging themes of Gli activities in epidermal homeostasis.

splitting Hairs

103

I u

U

u ^ ^

Q. CO C ^

U c _:

"D

vX5

c ,

o o

(^ _Q

QC

Q - CM

t

c o

i>5 • > -

3

^ ^ T—

C

•^ T3 fl)

^

In -^

-B

o _y

n3

X

m

^"^ I ^"^^

X

F cu £ S^ o 2 o o

03

03

03

1

>

03

CD

•D

c DO C

CD

CD 3 ^ ^

"a c

^ ^^

u c

Sf - D

CD

X

^ Q

to

CD Q . CD " D C

to

03

to

to

to

1-

CD

c

a;

< u

< u -a

U CD Q . i>j

CD to C O Q.

U

2f D S^

3

o

C CD DO DO "O D

DO

c ^ O i l O 03 Q. U Q. DO to 03 to QJ to _ I CD _DJ O ^ I

I

c D

C

0)

CD 03

_C

-Si

2

O

O

u

u

CD to DO OS O CD

-^ ^

.. O o

< u 1 2-8

U

^ ^ s ."ti (D O CD

c

D "t: S

i

3 DO

CD 3 DO DO O CD

c

II

.CN

Q

CN

S &^

I—

CN

i

CN

X y^ I

to

CO

03 C 1 .DO ( J ' t o •

^1^

^ ICyc/m Bi

CycttnDI

hyciinm

• ceacveLg PROORiSSION

SE RESTRICTION POINT

Figure 7. A model for Gli mediated Shh target gene expression controlling proliferation in developing hair follicle. Shh signaling controls cell cycle progression in the developing epidermis through transcriptional regulation of multiple cell cycle regulators. Although Shh''; GliS^' and GliZ' follicles are partially rescued with respect to the expression of early cell cycle regulators, including N-myc and CyclinD2y they both lack Gli activator function that is essential for cell cycle progression through transcriptional control and post-transcriptional stabilization of later cell cycle regulators. N-myc protein is not stabilized in Shh''; G//3'' skin, suggesting that there are additional levels of control that require Shh-dependent Gli activator function. This result is consistent with a recent report that proliferation in cerebellar neuronal precursor cells requires both Shh-mediated transcriptional activation and phosphatidyiinositol-3-kinase (PI-3K) mediated inhibition of N-myc degradation.^'^ Furthermore, we found robust proliferation requires progenitors to move concertedly through the cell cycle, which involves the switch from repressor to activator E2Fs to activate Gi/S specific genes at the restriction point. This progression depends on sustained expression and/or their accumulation which additionally requires Shh-dependent Gli activator function (P.M. and C.c.H. manuscript in preparation). During embryonic and adult hair follicle development, Shh induces proliferation of hair follicle progenitors through induction of a transcriptional hierarchy of cell cycle regulators, including N-myc, CyclinDl, CyclinD2 and E2Fs. In contrast, hair follicles lacking Shh do not express these mitogenic markers, as is the case with arrested Shh mutant hair follicles and adult telogen follicles, which show no significant proliferation (RM. and C.c.Fi. unpublished results). Similar expression profiles of cell cycle target genes to those observed in anagen hair follicles are detected in K5Cre; Z/APGli2 and K5Cre; Z/APANGli2 tumors, where the level of expression of these genes correlates with the tumor proliferative index. It was recently reported that overexpression of Gli2 induces rapidly expression of similar mitogenic targets, including CyclinDl and E2F1, sufficiently to override growth arrest in contact inhibited keratinocyte cultures. ^^^ Hyperproliferative changes were observed early in K5Cre; Z/APGli2 and K5Cre; Zl APANGli2 skin prior to induction of tumors, including upregulated expression oiE2Fl (P.M. and C.cH. unpublished results). All three epidermal lineages appeared to be sensitive to hyperactive Hh responses as shown by thickening of the interfoUicular epidermis, sebaceous gland enlargement and alterations in hair follicle orientations. Fiowever, only hair follicle-derived tumors develop in K5Cre; Z/APGli2 and K5Cre; Z/APANGli2 skin. These results suggest that deregtdated Hh signaling can act as a potent mitogen for multiple epidermal lineages, but its transforming ability is limited to progenitors destined for hair follicle fates. Similar Shh-dependent Gli activator and repressor functions likely regulate proliferation at multiple cell cycle checkpoints in adult skin, where Shh proliferative responses are tightly controlled during hair cycle and deregulated during tumorigenesis. Furthermore, given that

Spliuing Hairs

109

Shh leads to induction and stabilization of N-myc associated with increased proliferation of granule cell precursors and medulloblastomas,^^ ' our results could provide a proximal "missing link" to these processes through opposing activities of the Gli transcription factors. The epidermis is an ideal developmental system to address these issues, as Shh plays primarily a proliferative role (RM. and C.cH. manuscript in preparation).^'^^'^^ '^^^ Understanding the transcription and stabilization profile of target cell cycle regulators controlled by Gli activator and repressor functions will provide important insights into many developmental and disease states, as well as potential therapeutic targets for treatment of pathological conditions where Hh signaling is deregulated.

Conclusions Using a genetic approach complementing loss-of-function and gain-of-fiinction mouse mutants, we have been able to dissect temporal and spatial requirements for Shh signaling in the skin. During hair follicle morphogenesis, Shh signaling is primarily required by the epidermis to promote proliferation of epidermal progenitors. Two important functions of Shh signaling are required for mitogenic responses to occur; (1) it promotes the generation of Gli activators, especially Gli2 and (2) it inhibits the generation of Gli repressors, primarily Gli3 Interpreted by the cell as a "Gli code", the summation of these Gli activities determines the transcriptional activity of target genes and ultimately, the decision to enter the cell cycle and proliferate. These studies have revealed novel Gli-dependent regulation of a transcriptional hierarchy of key cell cycle regulators, including the D-type cyclinSy N-myc and E2Fs. Moreover, it has unveiled functional differences in how these target genes are regulated by Gli proteins. Expression of early cell cycle markers was derepressed in Shh; Gli3 double mutants, but induaion of later regulators, namely activator E2Fsy could not be induced without Gli"^ . Consequendy, robust proliferative responses were not observed. These results suggest that Shh signaling controls the cell cycle at multiple levels, to ensure that proliferative responses are restricted in time and space to regions of high Hh signal. The Hh-dependent balance of Gli activator and repressor functions also plays a key role in limiting proliferative responses to adult hair follicles at the telogen-anagen transition. Disturbing this balance, either by increasing the magnitude of Hh responses transiently at anagen in K5Cre; Z/APGli2 mice or by extending Hh responses independently of the hair cycle in K5Cre; Z/APANGli2 mice, is sufficient to overwhelm endogenous regulation. These changes disrupt epidermal homeostasis, triggering hyperproliferative responses and tumors in transgenic skin. There appear to be distinct classes of target genes that are differentially regulated by Gli activator and repressor functions (Fig. 1). Mutant analysis has revealed that certain genes are induced by Gli2, while expression of other targets is inhibited by Gli3 in the embryonic skin. As seen in the developing neural tube, expression ofPtcl, but not Glily is derepressed in Shh; Gli3 vibrissae and pelage follicles. The expression o£ N-myc and cyclinD2 is also derepressed in the developing hair follicles by removing Gli3 repressor activity in Shh mutants. Since expression of Ptcl, N-myCy and cyclinD2 can also be induced by Gli activator function (P.M. an C.c.H. manuscript in preparation),^^ these genes represent a class of Shh targets that are regulated by both Gli activator and repressor. In contrast, the transcription of other Shh target genes, such as GUI and Wnt5ay seem to be regulated solely by Gli activator function. Furthermore, it is possible that there is a third class of Shh target genes, which are regulated solely by Gli repressor. Shh can induce target gene transcription through promoting the Gli activator function and/by inhibiting the Gli repressor function. Our genetic studies in skin have revealed a Gli-dependent transcriptional cascade of cell cycle regulators in vivo, the next step will be to determine how these targets are directly or indirectly regulated by Gli activities. An interesting aspect of the Hh intracellular response is the parallel induction of a positive transcriptional feedback loop through GUI, which propagates and amplifies Hh transcriptional responses, and a negative signaling feedback loop through Ptcl and Hipl (Hedgehog binding protein) which serve to dampen Hh signaling at the cell surface (Fig. 8).


110 BC€ {HM)

BC€(H,M)

Negative

sec {HM) Trichoblastoma (H?,M) 'Trichoepitlieiioma (H) Sefeaoeous nevys (H?) BCC (H,M) Basaloid hamartoma (M) Selâceous nevys (H) THchoblastoma (M)

.^H

regulation

.e Ptd. Hlpi)// 8CC (M) Basaloid hamartoma (M) Trichoblastoma (M) Trichoeptthaiioma (M) \

/.a G//7 tProliferation /.a cyclfnDI^ tCell survival /.a ibc/-2 ^Differentiation/.a K/7

Tumor initiation & maintenance

Figure 8. Hedgehog signaling and skin cancer. Alterations in various components of the Hh signaling cascade can result in both human (H) and mouse (M) skin tumors, by either inaaivating negative regulators (i.e., Ptc) or deregulating positive factors (I.e., Smo). Such conditions result in the accumulation of Gli activator species, which drive expression of cell cycle regulators (i.e., D-type cyclins)y survival factors (i.e., bcl'2) and stem-cell like characteristics (i.e., K17) required to support tumor initiation and maintenance. Hh signals induce expression of another set of target genes, including Ptcl and Hipl, which compose a negative feedback loop at the distal end of the pathway. However, once a tumor is established, it has become ligand-independent through the autoregulation of GUI and leads to uncontrolled proliferation. T h e interplay between these feedback loops create a "genetic switch", whereby a specific developmental decision is made, such as choice of cell fate, at a distinct threshold of Shh signal."^^^ In disease settings where the negative reetilatory loop is compromised, such as reported truncated or dominant-negative Ptc mutations, ^^' ^ the system may be incapable of turning itself "OFF" resulting in unregulated H h responses and unchecked proliferation. Accordingly, high levels of negative feedback components Hipl and Ptcl expression are detected in mouse and human BCCs.^^^ Alternately, increasing Gli activator species either artificially in transgenic mouse models (P.M. and C.c.H., manuscript in preparation)^ ^'^ or by unclear mechanisms in human skin^^^ may overwhelm the endogenous negative feedback loop in a ligand-independent hyperactivation of H h responses and proliferation. Consideration of these regulatory loops needs be incorporated into future therapeutic strategies for the prevention and treatment of Hh-related cancers. Initial reports of upstream antagonists, such as those opposing Smoothened fiinctions including cyclopamine and small molecule inhibitor CUR161414, show effective BCC lesion regression with minimal side effects in animal and clinical trials. However, it is possible that some Hh-dependent tumors, such as those resulting from overexpression of Gli activator species in mice or loss of proximal negative regulator Su(Fu) in some human meduUoblastoma patients,^ will be refractory to these distal treatments that have litde impact on limiting the hyperactive positive Gli feedback loop.^^^ An effective alternate strategy could be to target this feedback loop direcdy, such as with GUI R N A interference. ^^^ If the balance of Gli activity is ultimately the deciding factor in developmental and disease choices within a cell, better understanding of the molecular mechanisms which determine this set point of Gli activator and repressor levels needs to be the focus of future studies.

splitting

Hairs

111

References 1. Chiang C, Swan RZ, Grachtchouk M et al. Essential role for Sonic hedgehog during hair follicle morphogenesis. Dev Biol 1999; 205(1): 1-9. 2. Sato N , Leopold PL, Crystal RG. Induction of the hair growth phase in postnatal mice by localized transient expression of Sonic hedgehog. J CUn Invest 1999; 104(7):855-864. 3. St-Jacques B, Dassule HR, Karavanova I et al. Sonic hedgehog signaling is essential for hair development. Curr Biol 1998; 8(19):1058-1068. 4. W a n g LC, Liu ZY, Gambardella L et al. Regular articles: Conditional disruption of hedgehog signaling pathway defines its critical role in hair development and regeneration. J Invest Dermatol 2000; l l 4 ( 5 ) : 9 0 1 - 9 0 8 . 5. Aszterbaum M , Epstein J, O r o A et al. Ultraviolet and ionizing radiation enhance the growth of B C C s a n d t r i c h o b l a s t o m a s in p a t c h e d h e t e r o z y g o u s k n o c k o u t m i c e . N a t M e d 1 9 9 9 ; 5(11):1285-1291. 6. Aszterbaum M , Rothman A, Johnson RL et al. Identification of mutations in the human P A T C H E D gene in sporadic basal cell carcinomas and in patients with the basal cell nevus syndrome. J Invest Dermatol 1998; 110(6):885-888. 7. Dahmane N , Lee J, Robins P et al. Activation of the transcription factor GUI and the Sonic hedgehog signalling pathway in skin tumours. Nature 1997; 389(6653):876-881. 8. H a h n H , Wicking C, Zaphiropoulous PG et al. Mutations of the human homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome. Cell 1996; 8 5 ( 6 ) : 8 4 l - 8 5 1 . 9. Johnson RL, Rothman AL, Xie J et al. H u m a n homolog of patched, a candidate gene for the basal cell nevus syndrome. Science 1996; 272 (5268): 1668-1671. 10. Ingham P W , M c M a h o n AP. Hedgehog signaling in animal development: Paradigms and principles. Genes Dev 2 0 0 1 ; 15(23):3059-3087. 11. Alexandre C, Jacinto A, Ingham PW. Transcriptional activation of hedgehog target genes in Drosophila is mediated directly by the cubitus interruptus protein, a member of the GLI family of zinc finger DNA-binding proteins. Genes Dev 1996; 10(16):2003-2013. 12. Methot N , Basler K. An absolute requirement for Cubitus interruptus in Hedgehog signaling. Development 2 0 0 1 ; 128(5):733-742. 13. Ruiz i Altaba A, Nguyen V, Palma V. T h e emergent design of the neural tube: Prepattern, S H H morphogen and GLI code. Curr O p i n Genet Dev 2003; 13(5):513-521. 14. Von Ohlen T , Lessing D , Nusse R et al. Hedgehog signaling regulates transcription through cubitus interruptus, a sequence-specific D N A b i n d i n g protein. Proc N a t l Acad Sci USA 1997; 94(6):2404-2409. 15. Chen Y, Struhl G. In vivo evidence that Patched and Smoothened constitute distinct binding and transducing components of a Hedgehog receptor complex. Development 1998; 125(24):4943-4948. 16. Stegman MA, Vallance JE, Elangovan G et al. Identification of a tetrameric hedgehog signaling complex. J Biol Chem 2000; 275(29):21809-21812. 17. W a n g G, Jiang J. Multiple Cos2/Ci interactions regulate Ci subcellular localization through microtubule dependent and independent mechanisms. Dev Biol 2004; 268(2):493-505. 18. Zhang W , Zhao Y, T o n g C et al. Hedgehog-regulated Costal2-kinase complexes control phosphorylation and proteolytic processing of Cubitus interruptus. Dev Cell 2005; 8(2):267-278. 19. Chen Y, Cardinaux JR, Goodman R H et al. Mutants of cubitus interruptus that are independent of PKA regulation are independent of hedgehog signaling. Development 1999; 126(16):3607-3616. 20. Price MA, Kalderon D . Proteolysis of cubitus interruptus in Drosophila requires phosphorylation by protein kinase A. Development 1999; 126(19):4331-4339. 2 1 . W a n g G, W a n g B, Jiang J. Protein kinase A antagonizes Hedgehog signaling by regulating both the activator and repressor forms of Cubitus interruptus. Genes Dev 1999; 13(21)-.2828-2837. 22. Aza-Blanc P, Ramirez-Weber FA, Laget M P et al. Proteolysis that is inhibited by hedgehog targets C u b i t u s i n t e r r u p t u s p r o t e i n to t h e n u c l e u s a n d converts it to a repressor. Cell 1 9 9 7 ; 89(7):1043-1053. 23. Jiang J, Struhl G. Regulation of the Hedgehog and Wingless signalling pathways by the F-box/ WD40-repeat protein Slimb. Nature 1998; 391 (6666):493-496. 24. Methot N , Basler K. Hedgehog controls limb development by regulating the activities of distinct transcriptional activator and repressor forms of Cubitus interruptus. Cell 1999; 96(6):819-831. 25. Methot N , Basler K. Suppressor of fused opposes hedgehog signal transduction by impeding nuclear accumulation of the activator form of Cubitus interruptus. Development 2000; 127(18):4001-4010. 26. Ohlmeyer J T , Kalderon D . Hedgehog stimulates maturation of Cubitus interruptus into a labile transcriptional activator. Nature 1998; 396(6713):749-753. 27. W a n g Q T , Holmgren RA. Nuclear import of cubitus interruptus is regulated by hedgehog via a mechanism distinct from Ci stabilization and Ci activation. Development 2000; 127(14):3131-3139.

112

Hedgehog-Gli

Signaling in Human

Disease

28. Hooper JE. Smoothened translates Hedgehog levels into distinct responses. Development 2003; 130(17):3951-3963. 29. Lum L, Zhang C, O h S et al. Hedgehog signal transduction via Smoothened association with a cytoplasmic complex scaffolded by the atypical kinesin, Costal-2. Mol Cell 2003; 12(5): 1261-1274. 30. Ogden SK, Ascano Jr M , Stegman MA et al. Identification of a functional interaction between the transmembrane protein Smoothened and the kinesin-related protein Costal2. Curr Biol 2003; 13(22): 1998-2003. 3 1 . Ruel L, Rodriguez R, Gallet A et al. Stability and association of Smoothened, Costal2 and Fused with Cubitus interruptus are regulated by Hedgehog. Nat Cell Biol 2003; 5(10):907-913. 32. Z h u AJ, Zheng L, Suyama K et al. Altered localization of Drosophila Smoothened protein activates Hedgehog signal transduction. Genes Dev 2003; 17(10):1240-1252. 33. Jia J, T o n g C, Jiang J. Smoothened transduces Hedgehog signal by physically interacting with Costal2/Fused complex through its C-terminal tail. Genes Dev 2003; 17(21):2709-2720. 34. Pathi S, Pagan-Westphal S, Baker D P et al. Comparative biological responses to human Sonic, Indian, and Desert hedgehog. Mech Dev 2 0 0 1 ; 106(1-2):107-117. 35. Carpenter D , Stone D M , Brush J et al. Characterization of two patched receptors for the vertebrate hedgehog protein family. Proc Natl Acad Sci USA 1998; 95(23): 13630-13634. 36. Motoyama J, Takabatake T , Takeshima K et al. Ptch2, a second mouse Patched gene is coexpressed with Sonic hedgehog. N a t Genet 1998; 18(2):104-106. 37. Pearse Ilnd RV, Vogan KJ, Tabin CJ. Ptcl and Ptc2 transcripts provide distinct readouts of Hedgehog signaling activity during chick embryogenesis. Dev Biol 2 0 0 1 ; 239(l):15-29. 38. Rahnama F, Toftgard R, Zaphiropoulos PG. Distinct roles of P T C H 2 splice variants in Hedgehog signalling. Biochem J 2004; 378(Pt 2):325-334. 39. Taipale J, Cooper MK, Maiti T et al. Patched acts catalytically to suppress the activity of Smoothened. Nature 2002; 4l8(6900):892-897. 40. Murone M, Rosenthal A, de Sauvage FJ. Sonic hedgehog signaling by the patched-smoothened receptor complex. Curr Biol 1999; 9(2):76-84. 4 1 . Ding Q , Fukami S, M e n g X et al. Mouse suppressor of fused is a negative regulator of sonic hedgehog signaling and alters the subcellular distribution of GUI. Curr Biol 1999; 9(19):1119-1122. 42. Liu A, W a n g B, Niswander LA. Mouse intraflagellar transport proteins regulate both the activator and repressor functions of Gli transcription factors. Development 2005; 132(13):3103-3111. 4 3 . Murone M , Luoh SM, Stone D et al. Gli regulation by the opposing activities of fused and suppressor of fused. Nat Cell Biol 2000; 2(5):310-312. 44. Kim J, Kim P, Hui C C . T h e VACTERL association: Lessons from the Sonic hedgehog pathway. Clin Genet 2 0 0 1 ; 59(5):306-315. 45. Matise M P , Epstein DJ, Park H L et al. Gli2 is required for induction of floor plate and adjacent cells, but not most ventral neurons in the mouse central nervous system. Development 1998; 125(15):2759-2770. 46. Lee J, Piatt KA, Censullo P et al. Glil is a target of Sonic hedgehog that induces ventral neural tube development. Development 1997; 124(13):2537-2552. 47. Marigo V, Johnson RL, Vortkamp A et al. Sonic hedgehog differentially regulates expression of GLI and GLI3 during limb development. Dev Biol 1996; 180(l):273-283. 48. Matise M P , Joyner AL. Gli genes in development and cancer. Oncogene 1999; 18(55):7852-7859. 49. Sasaki H , Nishizaki Y, H u i C et al. Regulation of Gli2 and Gli3 activities by an amino-terminal repression domain: Implication of Gli2 and Gli3 as primary mediators of Shh signaling. Development 1999; 126(17):3915-3924. 50. Dai P, Akimaru H , Tanaka Y et al. Sonic Hedgehog-induced activation of the Glil promoter is mediated by GLI3. J Biol Chem 1999; 274(12):8143-8152. 5 1 . Ruiz i Altaba A. Gli proteins encode context-dependent positive and negative functions: Implications for development and disease. Development 1999; 126(14):3205-3216. 52. Shin SH, Kogerman P, Lindstrom E et al. GLI3 mutations in human disorders mimic Drosophila c u b i t u s i n t e r r u p t u s p r o t e i n functions and localization. Proc N a t l Acad Sci USA 1999; 96(6):2880-2884. 53. Lei Q , Zelman AK, Kuang E et al. Transduction of graded Hedgehog signaling by a combination of Gli2 and Gli3 activator functions in the developing spinal cord. D e v e l o p m e n t 2 0 0 4 ; 131(15):3593-3604. 54. Stamataki D , Ulloa F, Tsoni SV et al. A gradient of Gli activity mediates graded Sonic Hedgehog signaling in the neural tube. Genes Dev 2005; 19(5):626-64l. 55. Aza-Blanc P, Lin HY, Ruiz i Altaba A et al. Expression of the vertebrate Gli proteins in Drosop h i l a reveals a d i s t r i b u t i o n of a c t i v a t o r a n d repressor a c t i v i t i e s . D e v e l o p m e n t 2 0 0 0 ; 127(19):4293-4301.

splitting

Hairs

113

56. von Mering C, Basler K. Distinct and regulated activities of human Gli proteins in Drosophila. Curr Biol 1999; 9(22):1319-1322. 57. Bai C B , Joyner AL. G l i l can rescue the in vivo function of G l i 2 . D e v e l o p m e n t 2 0 0 1 ; 128(24):5161-5172. 58. Bai CB, Stephen D , Joyner AL. All mouse ventral spinal cord patterning by hedgehog is Gli dependent and involves an activator function of Gli3. Dev Cell 2004; 6(1): 103-115. 59. Buttitta L, M o R, H u i C C et al. Interplays of Gli2 and Gli3 and their requirement in mediating Shh-dependent sclerotome induction. Development 2003; 130(25):6233-6243. 60. Motoyama J, Milenkovic L, Iwama M et al. Differential requirement for Gli2 and Gli3 in ventral neural cell fate specification. Dev Biol 2003; 259(1):150-161. 6 1 . Li Y, Zhang H , Choi SC et al. Sonic hedgehog signaling regulates Gli3 processing, mesenchymal proliferation, and differentiation during mouse lung organogenesis. Dev Biol 2004; 270(1):214-231. 62. MuUer B, Basler K. The repressor and activator forms of Cubitus interruptus control Hedgehog target genes through common generic gli-binding sites. Development 2000; 127(14):2999-3007. 63. Bai CB, Auerbach W , Lee JS et al. Gli2, but not G l i l , is required for initial Shh signaling and ectopic activation of the Shh pathway. Development 2002; 129(20):4753-4761. GA. Park HL, Bai C, Piatt KA et al. Mouse Glil mutants are viable but have defects in S H H signaling in combination with a Gli2 mutation. Development 2000; 127(8): 1593-1605. 65. Ding Q , Motoyama J, Gasca S et al. Diminished Sonic hedgehog signaling and lack of floor plate differentiation in Gli2 mutant mice. Development 1998; 125(l4):2533-2543. G6. M o R, Freer A M , Zinyk D L et al. Specific and redundant functions of Gli2 and Gli3 zinc finger genes in skeletal patterning and development. Development 1997; 124(1): 113-123. G7. Motoyama J, Liu J, M o R et al. Essential function of Gli2 and Gli3 in the formation of lung, trachea and o e s o p h ^ u s . Nat Genet 1998; 20(l):54-57. 68. H u i C C , Joyner AL. A mouse model of greig cephalopolysyndactyly syndrome: T h e extra-toesj mutation contains an intragenic deletion of the GU3 gene. N a t Genet 1993; 3(3):24l-246. 69. Litingtung Y, Chiang C. Specification of ventral neuron types is mediated by an antagonistic interaction between Shh and GH3. N a t Neurosci 2000; 3(10):979-985. 70. Schimmang T , Lemaistre M , Vortkamp A et al. Expression of the zinc finger gene Gli3 is affected in the morphogenetic mouse mutant extra-toes (Xt). Development 1992; 116(3):799-804. 7 1 . Aoto K, Nishimura T , Eto K et al. Mouse GLI3 regulates Fgf8 expression and apoptosis in the developing neural tube, face, and limb bud. Dev Biol 2002; 251(2):320-332. 72. Litingtung Y, D a h n R D , Li Y et al. Shh and Gli3 are dispensable for limb skeleton formation but regulate digit number and identity. Nature 2002; 418(6901):979-983. 7 3 . Persson M , Stamataki D , te Welscher P et al. Dorsal-ventral patterning of the spinal cord requires Gli3 transcriptional repressor activity. Genes Dev 2002; 16(22):2865-2878. lA. te Welscher P, Zuniga A, Kuijper S et al. Progression of vertebrate limb development through SHH-mediated counteraction of GLI3. Science 2002; 298(5594):827-830. 75. Gallet A, Angelats C, Kerridge S et al. Cubitus interruptus-independent transduction of the Hedgehog signal in Drosophila. Development 2000; 127(24):5509-5522. 7G. Krishnan V, Pereira FA, Q i u Y et al. Mediation of Sonic hedgehog-induced expression of C O U P - T F I I by a protein phosphatase. Science 1997; 278(5345): 1947-1950. 77. Lessing D , Nusse R. Expression of wingless in the Drosophila embryo: A conserved cis-acting element lacking conserved Ci-binding sites is required for patched-mediated repression. Development 1998; 125(8):1469-1476. 7 8 . Suzuki T , Saigo K. Transcriptional regulation of atonal required for Drosophila larval eye development by concerted action of eyes absent, sine oculis and hedgehog signaling independent of fused kinase and cubitus interruptus. Development 2000; 127(7):1531-1540. 79. Testaz S, Jarov A, Williams KP et al. Sonic hedgehog restricts adhesion and migration of neural crest cells independently of the Patched- Smoothened-Gli signaling pathway. Proc Natl Acad Sci USA 2 0 0 1 ; 98(22):12521-12526. 80. H u i C C , Slusarski D , Piatt KA et al. Expression of three mouse homologs of the Drosophila segment polarity gene cubitus interruptus, Gli, Gli-2, and Gli-3, in ectoderm- and mesoderm-derived tissues suggests multiple roles during postimplantation development. Dev Biol 1994; 162(2):402-413. 8 1 . Brewster R, MuUor JL, Ruiz i Altaba A. Gli2 functions in FGF signaling during antero-posterior patterning. Development 2000; 127(20):4395-4405. 82. Liu F, Massague J, Ruiz i Altaba A. Carboxy-terminally truncated Gli3 proteins associate with Smads. N a t Genet 1998; 20(4):325-326. 83. Millar SE. Molecular mechanisms regulating hair folUcle development. J Invest Dermatol 2002; 118(2):216-225. 84. Hardy M H . T h e secret life of the hair follicle. Trends Genet 1992; 8(2):55-61.

114

Hedgehog-Gli

Signaling in Human

Disease

85. Fuchs E, Merrill BJ, Jamora C et al. At the roots of a never-ending cycle. Dev Cell 2001; l(l):13-25. 86. Lavker RM, Sun T T , Oshima H et al. Hair follicle stem cells. J Investig Dermatol Symp Proc 2003; 8(l):28-38. 87. Muller-Rover S, Handjiski B, van der Veen C et al. A comprehensive guide for the accurate classification of murine hair follicles in distinct hair cycle stages. J Invest Dermatol 2 0 0 1 ; 117(1):3-15. 88. Jahoda CA. Cell movement in the hair follicle dermis - More than a two-way street? J Invest Dermatol 2003; 121(6):ix-xi. 89. Niemann C, W a t t F M . Designer skin: Lineage commitment in postnatal epidermis. Trends Cell Biol 2002; 12(4):185-192. 90. Botchkarev VA, Kishimoto J. Molecular control of epithelial-mesenchymal interactions during hair follicle cycling. J Investig Dermatol Symp Proc 2003; 8(l):46-55. 9 1 . Tumbar T , Guasch G, Greco V et al. Defining the epithelial stem cell niche in skin. Science 2004; 303(5656):359-363. 92. Bitgood MJ, McMahon AP. Hedgehog and Bmp genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev Biol 1995; 172(1): 126-138. 9 3 . Iseki S, Araga A, Ohuchi H et al. Sonic hedgehog is expressed in epithelial cells during development of whisker, hair, and tooth. Biochem Biophys Res C o m m u n 1996; 218(3):688-693. 94. Morgan BA, Orkin RW, Noramly S et al. Stage-specific effects of sonic hedgehog expression in the epidermis. Dev Biol 1998; 201(1):1-12. 9 5 . Ting-Berreth SA, C h u o n g C M . Sonic Hedgehog in feather morphogenesis: Induction of mesenchymal condensation and association with cell death. Dev Dyn 1996; 207(2): 157-170. 96. O r o AE, Higgins KM, H u Z et al. Basal cell carcinomas in mice overexpressing sonic hedgehog. Science 1997; 276(5313):817-821. 97. Karlsson L, Bondjers C, Betsholtz C. Roles for PDGF-A and sonic hedgehog in development of mesenchymal components of the hair follicle. Development 1999; 126(12):2611-2621. 98. Reddy S, Andl T , Bsâsra A et al. Characterization of W n t gene expression in developing and postnatal hair follicles and identification of W n t 5 a as a target of Sonic hedgehog in hair follicle morphogenesis. Mech Dev 2 0 0 1 ; 107(l-2):69-82. 99. Mill P, M o R, Fu H et al. Sonic hedgehog-dependent activation of Gli2 is essential for embryonic hair foHicle development. Genes Dev 2003; 17(2):282-294. 100. O r o AE, Higgins K. Hair cycle regulation of H e d g e h o g signal reception. Dev Biol 2 0 0 3 ; 255(2):238-248. 101. Allen M, Grachtchouk M , Sheng H et al. Hedgehog signaling regulates sebaceous gland development. Am J Pathol 2003; 163(6):2173-2178. 102. Niemann C, Unden AB, Lyle S et al. Indian hedgehog and beta-catenin signaling: Role in the sebaceous lineage of normal and neoplastic mammalian epidermis. Proc Natl Acad Sci USA 2003; 100(Suppl 1):11873-11880. 103. Wijgerde M , McMahon JA, Rule M et al. A direct requirement for Hedgehog signaling for normal specification of all ventral progenitor domains in the presumptive mammalian spinal cord. Genes Dev 2002; 16(22):2849-2864. 104. Meyer N P , Roelink H . T h e amino-terminal region of Gli3 antagonizes the Shh response and acts in dorsoventral fate specification in the developing spinal cord. Dev Biol 2003; 257(2):343-355. 105. Nicolas M , Wolfer A, Raj K et al. N o t c h l functions as a tumor suppressor in mouse skin. Nat Genet 2003; 33(3):416-421. 106. Andl T, Reddy ST, Gaddapara T et al. W N T signals are required for the initiation of hair follicle development. Dev Cell 2002; 2(5):643-653. 107. Huelsken J, Vogel R, Erdmann B et al. beta-Catenin controls hair follicle morphogenesis and stem cell differentiation in the skin. Cell 2 0 0 1 ; 105(4):533-545. 108. Bonifas J M , Pennypacker S, Chuang P T et al. Activation of expression of hedgehog target genes in basal cell carcinomas. J Invest Dermatol 2 0 0 1 ; 116(5):739-742. 109. Lee CS, Buttitta L, Fan C M . Evidence that the WNT-inducible growth arrest-specific gene 1 encodes an antagonist of sonic hedgehog signaling in the somite. Proc Natl Acad Sci USA 2001; 98(20): 11347-11352. 110. Cobourne M T , Miletich I, Sharpe PT. Restriction of sonic hedgehog signalling during early tooth development. Development 2004; 131(12):2875-2885. 111. Lee CS, Fan C M . Embryonic expression patterns of the mouse and chick Gasl genes. Mech Dev 2001; 101(l-2):293-297. 112. Chuong C M , Patel N , Lin J et al. Sonic hedgehog signaling pathway in vertebrate epithelial appendage morphogenesis: Perspectives in development and evolution. Cell Mol Life Sci 2000; 57(12):1672-1681. 113. Ohsaki K, Osumi N , Nakamura S. Altered whisker patterns induced by ectopic expression of Shh are topographically represented by barrels. Brain Res Dev Brain Res 2002; 137(2): 159-170.

splitting

Hairs

115

114. Dassule H R , Lewis P, Bei M et al. Sonic hedgehog regulates growth and morphogenesis of the tooth. Development 2000; 127(22):4775-4785. 115. Jaskoll T , Leo T , Witcher D et al. Sonic hedgehog signaling plays an essential role during embryonic salivary gland epithelial branching morphogenesis. Dev Dyn 2004; 229(4):722-732. 116. Cobourne M T , Hardcastle Z, Sharpe P T . Sonic hedgehog regulates epitheUal proUferation and cell survival in the developing tooth germ. J Dent Res 2 0 0 1 ; 80(11):1974-1979. 117. Gritli-Linde A, Bei M, Maas R et al. Shh signaling within the dental epithelium is necessary for cell proUferation, growth and polarization. Development 2002; 129(23):5323-5337. 118. Hardcastle Z, H u i C C , Sharpe PT. T h e Shh signalling pathway in early tooth development. Cell Mol Biol (Noisy-le-grand) 1999; 45(5):567-578. 119. Hardcastle Z, M o R, H u i C C et al. T h e Shh signalUng pathway in tooth development: Defects in Gli2 and Gli3 mutants. Development 1998; 125(15):2803-2811. 120. Chiang C, Litingtung Y, Harris M P et al. Manifestation of the limb prepattern: Limb development in the absence of sonic hedgehog function. Dev Biol 2 0 0 1 ; 236(2):421-435. 121. Gallego M I , Beachy PA, Hennighausen L et al. Differential requirements for shh in mammary tissue and hair follicle morphogenesis. Dev Biol 2002; 249(1): 131-139. 122. Kraus P, Fraidenraich D , Loomis CA. Some distal limb structures develop in mice lacking Sonic hedgehog signaling. Mech Dev 2 0 0 1 ; 100(l):45-58. 123. Michno K, Boras-Granic K, Mill P et al. Shh expression is required for embryonic hair follicle but not mammary gland development. Dev Biol 2003; 264(1): 153-165. 124. Lewis M T , Ross S, Strickland PA et al. T h e Gli2 transcription factor is required for normal mouse mammary gland development. Dev Biol 2 0 0 1 ; 238(1): 133-144. 125. Lewis M T , Ross S, Strickland PA et al. Defects in mouse mammary gland development caused by conditional haploinsufficiency of Patched-1. Development 1999; 126(22):5181-5193. 126. Pasca di Magliano M , Hebrok M . Hedgehog signalling in cancer formation and maintenance. Nat Rev Cancer 2003; 3(12):903-911. 127. Ruiz i Altaba A, Sanchez P, Dahmane N . Gli and hedgehog in cancer: T u m o u r s , embryos and stem cells. N a t Rev Cancer 2002; 2(5):361-372. 128. Berman D M , Karhadkar SS, Hallahan AR et al. MeduUoblastoma growth inhibition by hedgehog pathway blockade. Science 2002; 297(5586):1559-156l. 129. Karhadkar SS, Bova GS, Abdallah N et al. Hedgehog signalling in prostate regeneration, neoplasia and metastasis. Nature 2004; 431(7009):707-712. 130. Sanchez P , H e r n a n d e z A M , Stecca B et al. I n h i b i t i o n of prostate cancer proliferation by interference w i t h S O N I C H E D G E H O G - G L I l signaling. Proc N a t l Acad Sci U S A 2 0 0 4 ; 101(34):12561-12566. 131. Sheng T, Li C, Zhang X et al. Activation of the hedgehog pathway in advanced prostate cancer. Mol Cancer 2004; 3(1):29. 132. Thayer SP, di Magliano M P , Heiser P W et al. Hedgehog is an early and late mediator of pancreatic cancer tumorigenesis. Nature 2003; 425(6960):851-856. 133. Watkins D N , Berman D M , Burkholder SG et al. Hedgehog signalling within airway epithelial progenitors and in small-cell lung cancer. Nature 2003; 422(6929):313-317. 134. Dahmane N , Sanchez P, Gitton Y et al. T h e Sonic Hedgehog-Gli pathway regulates dorsal brain growth and tumorigenesis. Development 2 0 0 1 ; 128(24):5201-5212. 135. Kayed H , Kleeff J, Keleg S et al. Indian hedgehog signaling pathway: Expression and regulation in pancreatic cancer. Int J Cancer 2004; 110(5):668-676. 136. Tabs S, Avci O . Induction of the differentiation and apoptosis of tumor cells in vivo with efficiency and selectivity. Eur J Dermatol 2004; 14(2):96-102. 137. Taipale J, Chen JK, Cooper M K et al. Effects of oncogenic mutations in Smoothened and Patched can be reversed by cyclopamine. Nature 2000; 406(6799): 1005-1009. 138. Williams JA, Guicherit O M , Zaharian BI et al. Identification of a small molecule inhibitor of the hedgehog signaling pathway: Effects on basal cell carcinoma-like lesions. Proc Natl Acad Sci USA 2003; 100(8):4616-4621. 139. Ghazizadeh S, Taichman LB. Multiple classes of stem cells in cutaneous epithelium: A lineage analysis of adult mouse skin. Embo J 2 0 0 1 ; 20(6):1215-1222. 140. Owens D M , W a t t FM. Contribution of stem cells and differentiated cells to epidermal tumours. N a t Rev Cancer 2003; 3(6):444-451. 141. Oshima H , Rochat A, Kedzia C et al. Morphogenesis and renewal of hair follicles from adult multipotent stem cells. Cell 2 0 0 1 ; 104(2):233-245. 142. Taylor G, Lehrer MS, Jensen PJ et al. Involvement of follicular stem cells in forming not only the follicle but also the epidermis. Cell 2000; 102(4):451-461. 143. Ellis T, Smyth I, Riley E et al. Overexpression of Sonic Hedgehog suppresses embryonic hair follicle morphogenesis. Dev Biol 2003; 263(2):203-215.

116

Hedgehog-Gli

Signaling in Human

Disease

144. Hutchin ME, Kariapper MS, Grachtchouk M et al. Sustained Hedgehog signaling is required for basal cell carcinoma proliferation and survival: Conditional skin tumorigenesis recapitulates the hair growth cycle. Genes Dev 2005; 19(2):2l4-223. 145. Grachtchouk M , M o R, Yu S et al. Basal cell carcinomas in mice overexpressing Gli2 in skin. Nat Genet 2000; 24(3):216-217. 146. Grachtchouk V, Grachtchouk M , Lowe L et al. T h e magnitude of hedgehog signaling activity defines skin tumor phenotype. Embo J 2003; 2 2 ( l l ) : 2 7 4 l - 2 7 5 1 . 147. Nilsson M, Unden AB, Krause D et al. Induction of basal cell carcinomas and trichoepitheliomas in mice overexpressing GLI-1. Proc Natl Acad Sci USA 2000; 97(7):3438-3443. 148. Xie J, Murone M , Luoh SM et al. Activating Smoothened mutations in sporadic basal-cell carcinoma. Nature 1998; 391(6662):90-92. 149. W o n g CS, Strange RC, Lear JT. Basal cell carcinoma. Bmj 2003; 327(7418):794-798. 150. Gailani MR, Stable-Backdahl M, Leffell DJ et al. T h e role of the human homologue of Drosophila patched in sporadic basal cell carcinomas. Nat Genet 1996; 14(1):78-81. 151. Fan H , Oro AE, Scott M P et al. Induction of basal cell carcinoma features in transgenic human skin expressing Sonic Hedgehog. N a t Med 1997; 3(7):788-792. 152. Regl G, Neill G W , Eichberger T et al. H u m a n GLI2 and G L I l are part of a positive feedback mechanism in Basal Cell Carcinoma. Oncogene 2002; 21(36):5529-5539. 153. Goodrich LV, Milenkovic L, Higgins KM et al. Altered neural cell fates and medulloblastoma in mouse patched mutants. Science 1997; 277(5329):1109-1113. 154. H a h n H , Wojnowski L, Zimmer A M et al. Rhabdomyosarcomas and radiation hypersensitivity in a mouse model of Gorlin syndrome. N a t Med 1998; 4(5):619-622. 155. Pietsch T , Waha A, Koch A et al. Medulloblastomas of the desmoplastic variant carry mutations of the human homologue of Drosophila patched. Cancer Res 1997; 5 7 ( l l ) : 2 0 8 5 - 2 0 8 8 . 156. Raffel C, Jenkins RB, Frederick L et al. Sporadic medulloblastomas contain P T C H mutations. Cancer Res 1997; 57(5):842-845. 157. Reifenberger J, Wolter M , Weber RG et al. Missense mutations in S M O H in sporadic basal cell carcinomas of the skin and primitive neuroectodermal tumors of the central nervous system. Cancer Res 1998; 58(9):1798-1803. 158. Berman D M , Karhadkar SS, Maitra A et al. Widespread requirement for Hedgehog ligand stimulation in growth of digestive tract tumours. Nature 2003; 425(6960):846-851. 159. McGowan KM, Coulombc PA. Onset of keratin 17 expression coincides with the definition of major epithelial lineages during skin development. J Cell Biol 1998; l43(2):469-486. 160. Pazzaglia S, Mancuso M , Atkinson MJ et al. High incidence of medulloblastoma following X-ray-irradiation of newborn Ptcl heterozygous mice. Oncogene 2002; 21(49):7580-7584. 161. Perez-Losada J, Balmain A. Stem-cell hierarchy in skin cancer. Nat Rev Cancer 2003; 3(6):434-443. 162. Sheng H , Goich S, W a n g A et al. Dissecting the oncogenic potential of Gli2: Deletion of an NH(2)-terminal fragment alters skin tumor phenotype. Cancer Res 2002; 62(18):5308-5316. 163. Bigelow RL, Chari N S , Unden AB et al. Transcriptional regulation of bcl-2 mediated by the sonic hedgehog signaling pathway through gli-1. J Biol Chem 2004; 279(2):1197-1205. 164. Louro I D , Bailey EC, Li X et al. Comparative gene expression profile analysis of GLI and c-MYC in an epitheUal model of malignant transformation. Cancer Res 2002; 62(20):5867-5873. 165. Ling G, Ahmadian A, Persson A et al. P A T C H E D and p53 gene alterations in sporadic and hereditary basal cell cancer. Oncogene 2 0 0 1 ; 20(53):7770-7778. 166. Xie J, Aszterbaum M , Zhang X et al. A role of PDGFRalpha in basal cell carcinoma proliferation. Proc Natl Acad Sci USA 2001; 98(16):9255-9259. 167. Pomeroy SL, Tamayo P, Gaasenbeek M et al. Prediction of central nervous system embryonal tumour outcome based on gene expression. Nature 2002; 415(6870):436-442. 168. Rao G, Pedone CA, Valle LD et al. Sonic hedgehog and insulin-like growth factor signaling synergize to induce medulloblastoma formation from nestin-expressing neural progenitors in mice. Oncogene 2004; 23(36):6156-6162. 169. Bennett WR, Crew T E , Slack JM et al. Structural-proliferative units and organ growth: Effects of insulin-like growth factor 2 on the growth of colon and skin. Development 2003; 130(6): 1079-1088. 170. Louvi A, Accili D , Efstratiadis A. Growth-promoting interaction of IGF-II with the insulin receptor during mouse embryonic development. Dev Biol 1997; 189(l):33-48. 171. Philpott M P , Sanders DA, Kealey T . Effects of insulin and insuUn-Hke growth factors on cultured human hair follicles: IGF-I at physiologic concentrations is an important regulator of hair follicle growth in vitro. J Invest Dermatol 1994; 102(6):857-86l. 172. Rudman SM, Philpott M P , Thomas GA et al. The role of IGF-I in human skin and its appendages: Morphogen as well as mitogen? J Invest Dermatol 1997; 109(6):770-777. 173. Jia J, Amanai K, W a n g G et al. Shaggy/GSK3 antagonizes Hedgehog signalling by regulating Cubitus interruptus. Nature 2002; 4l6(6880):548-552.

splitting

Hairs

117

174. Kenney A M , Widlund H R , Rowitch D H , Hedgehog and PI-3 kinase signaling converge on N m y c l to p r o m o t e cell cycle progression in cerebellar n e u r o n a l precursors. D e v e l o p m e n t 2 0 0 4 ; 131(l):217-228. 175. Lum L, Beachy PA. T h e Hedgehog response network: Sensors, switches, and routers. Science 2004; 304(5678):1755-1759. 176. Yu JC, Gutkind JS, Mahadevan D et al. Biological function of PDGF-induced PI-3 kinase activity: Its role in alpha P D G F receptor-mediated mitogenic signaling. J Cell Biol 1994; 127(2):479-487. 177. Bellusci S, Furuta Y, Rush M G et al. Involvement of Sonic hedgehog (Shh) in mouse embryonic lung growth and morphogenesis. Development 1997; 124(l):53-63. 178. Grindley J C , Bellusci S, Perkins D et al. Evidence for the involvement of the Gli gene family in embryonic mouse lung development. Dev Biol 1997; 188(2):337-348. 179. Duprez D , Fournier-Thibault C, Le Douarin N . Sonic Hedgehog induces proliferation of committed skeletal muscle cells in the chick Umb. Development 1998; 125(3):495-505. 180. Marcelle C, Ahlgren S, Bronner-Fraser M . In vivo regulation of somite differentiation and proliferation by Sonic Hedgehog. Dev Biol 1999; 2l4(2):277-287. 181. Litingtung Y, Lei L, Westphal H et al. Sonic hedgehog is essential to foregut development. N a t Genet 1998; 20(1):58-61. 182. Roberts DJ, Smith D M , Goff DJ et al. Epithelial-mesenchymal signaling during the regionalization of the chick gut. Development 1998; 125(15):2791-2801. 183. Dahmane N , Ruiz i Altaba A. Sonic hedgehog regulates the growth and patterning of the cerebellum. Development 1999; 126(14):3089-3100. 184. Wallace VA. Purkinje-cell-derived Sonic hedgehog regulates granule neuron precursor cell proliferation in the developing mouse cerebellum. Curr Biol 1999; 9(8):445-448. 185. Wechsler-Reya RJ, Scott M P . Control of neuronal precursor proliferation in the cerebellum by Sonic Hedgehog. Neuron 1999; 22(1):103-114. 186. Zhang Y, Kalderon D . Hedgehog acts as a somatic stem cell factor in the Drosophila ovary. N a ture 2 0 0 1 ; 4l0(6828):599-604. 187. Lai K, Kaspar BK, Gage F H et al. Sonic hedgehog regulates adult neural progenitor proliferation in vitro and in vivo. N a t Neurosci 2003; 6(l):21-27. 188. Machold R, Hayashi S, Rutlin M et al. Sonic hedgehog is required for progenitor cell maintenance in telencephalic stem cell niches. Neuron 2003; 39(6):937-950. 189. Palma V, Lim DA, Dahmane N et al. Sonic hedgehog controls stem cell behavior in the postnatal and adult brain. Development 2005; 132(2):335-344. 190. Palma V, Ruiz i Altaba A. Hedgehog-GLI signaling regulates the behavior of cells with stem cell properties in the developing neocortex. Development 2004; 131(2):337-345. 191. Bhardwaj G, Murdoch B, W u D et al. Sonic hedgehog induces the proliferation of primitive human hematopoietic cells via BMP regulation. N a t Immunol 2 0 0 1 ; 2(2):172-180. 192. Kenney A M , Rowitch D H . Sonic hedgehog promotes G ( l ) cyclin expression and sustained cell cycle progression in mammalian neuronal precursors. Mol Cell Biol 2000; 20(23):9055-9067. 193. Long F, Zhang X M , Karp S et al. Genetic manipulation of hedgehog signaling in the endochondral skeleton reveals a direct role in the regulation of chondrocyte proliferation. Development 2 0 0 1 ; 128(24):5099-5108. 194. Fan H , Khavari PA. Sonic hedgehog opposes epithelial cell cycle arrest. J Cell Biol 1999; l47(l):71-76. 195. Barnes EA, Kong M , Ollendorff V et al. Patched 1 interacts with cyclin B l to regulate cell cycle progression. Embo J 2 0 0 1 ; 20(9):22l4-2223. 196. Duman-Scheel M , W e n g L, Xin S et al. Hedgehog regulates cell growth and proliferation by inducing Cyclin D and Cyclin E. Nature 2002; 417(6886):299-304. 197. Regl G, Kasper M , Schnidar H et al. T h e zinc-finger transcription factor GLI2 antagonizes contact inhibition and differentiation of h u m a n epidermal cells. Oncogene 2004; 23(6): 1263-1274. 198. Oliver T G , Grasfeder LL, Carroll AL et al. Transcriptional profiling of the Sonic hedgehog response: A critical role for N - m y c in proliferation of neuronal precursors. Proc Natl Acad Sci USA 2003; 100(12):7331-7336. 199. M o R, Kim J H , Zhang J et al. Anorectal malformations caused by defects in sonic hedgehog signaling. Am J Pathol 2 0 0 1 ; 159(2):765-774. 200. Lai K, Robertson MJ, Schaifer D V . T h e sonic hedgehog signaling system as a bistable genetic switch. Biophys J 2004; 86(5):2748-2757. 2 0 1 . Brellier F, Marionnet C, Chevallier-Lagente O et al. Ultraviolet irradiation represses P A T C H E D gene transcription in human epidermal keratinocytes through an activator protein-1-dependent process. Cancer Res 2004; 64(8):2699-2704.

118

Hedgehog-Gli

Signaling in Human

Disease

202. Hime GR, Lada H, Fietz MJ et al. Functional analysis in Drosophila indicates that the N B C C S / P T C H l mutation G 5 0 9 V results in activation of smoothened through a dominant-negative mechanism. Dev Dyn 2004; 229(4):780-790. 2 0 3 . Smyth I, Wicking C, Wainwright B et al. T h e effects of splice site mutations in patients with naevoid basal cell carcinoma syndrome. H u m Genet 1998; 102(5):598-601. 204. Wicking C, Shanley S, Smyth I et al. Most germ-line mutations in the nevoid basal cell carcinoma syndrome lead to a premature termination of the P A T C H E D protein, and no genotype-phenotype correlations are evident. A m J H u m Genet 1997; 60(l):21-26. 2 0 5 . Ikram MS, Neill G W , Regl G et al. GLI2 is expressed in normal human epidermis and BCC and induces G L I l expression by binding to its promoter. J Invest Dermatol 2004; 122(6): 1503-1509. 206. Taylor M D , Zhang X, Liu L et al. Failure of a meduUoblastoma-derived mutant of SUFU to suppress W N T signaling. Oncogene 2004; 23(26):4577-4583. 207. Bose J, Grotewold L, Ruther U. Pallister-Hall syndrome phenotype in mice mutant for Gli3. H u m Mol Genet 2002; 11(9):1129-1135. 208. D u SJ, Dienhart M . Gli2 mediation of hedgehog signals in slow muscle induction in zebrafish. Differentiation 2 0 0 1 ; 67(3):84-91. 209. Grove EA, Tole S, Limon J et al. T h e hem of the embryonic cerebral cortex is defined by the expression of multiple W n t genes and is compromised in Gli3-deficient mice. Development 1998; 125(12):2315-2325. 210. Kalff-Suske M , Wild A, T o p p J et al. Point mutations throughout the GLI3 gene cause Greig cephalopolysyndactyly syndrome. H u m Mol Genet 1999; 8(9): 1769-1777. 2 1 1 . Kang S, Graham Jr J M , Olney A H et al. GLI3 frameshift mutations cause autosomal dominant Pallister-Hall syndrome. N a t Genet 1997; 15(3):266-268. 212. Karlstrom R O , Tyurina O V , Kawakami A et al. Genetic analysis of zebrafish glil and gli2 reveals d i v e r g e n t r e q u i r e m e n t s for gli genes in v e r t e b r a t e d e v e l o p m e n t . D e v e l o p m e n t 2 0 0 3 ; 130(8):1549-1564. 2 1 3 . Kimmel SGy M o R, H u i C C et al. New mouse models of congenital anorectal malformations. J Pediatr Surg 2000; 35(2):227-230, (discussion 230-221). 214. Kondoh H , Uchikawa M , Yoda H ct al. Zebrafish mutations in Gli-mediated hedgehog signaling lead to lens transdifferentiation from the adenohypophysis anlage. Mcch Dev 2000; 96(2): 165-174. 215. Kuschel S, Ruther U, Theil T. A disrupted balance between B m p / W n t and Fgf signaling underlies the ventralization of the Gli3 mutant telencephalon. Dev Biol 2003; 260(2):484-495. 216. Qi Y, Tan M, Hui C C et al. Gli2 is required for normal Shh signaling and oligodendrocyte development in the spinal cord. Mol Cell Neurosci 2003; 23(3):440-450. 217. Radhakrishna U, Wild A, Grzeschik KH et al. Mutation in GLI3 in postaxial Polydactyly type A. N a t Genet 1997; 17(3):269-271. 218. Roessler E, D u YZ, Mullor JL et al. Loss-of-function mutations in the human GLI2 gene are associated with pituitary anomalies and holoprosencephaly-like features. Proc Natl Acad Sci USA 2003; 100(23):13424-13429. 219. Sbrogna JL, Barresi MJ, Karlstrom R O . Multiple roles for Hedgehog signaling in zebrafish pituitary development. Dev Biol 2003; 254(1): 19-35. 220. Tole S, Ragsdale C W , Grove EA. Dorsoventral patterning of the telencephalon is disrupted in the mouse mutant extra-toesQ). Dev Biol 2000; 217(2):254-265. 2 2 1 . Tyurina O V , Guner B, Popova E et al. Zebrafish Gli3 functions as both an activator and a repressor in Hedgehog signaling. Dev Biol 2005; 277(2):537-556. 222. Vanderlaan G, Tyurina O V , Karlstrom R O et al. Gli function is essential for motor neuron induction in zebrafish. Dev Biol 2005; 282(2):550-570. 2 2 3 . Vortkamp A, Gessler M , Grzeschik K H . GLI3 zinc-finger gene interrupted by translocations in Greig syndrome families. Nature 1991; 352(6335):539-540. 224. Wild A, Kalff-Suske M, Vortkamp A et al. Point mutations in human GLI3 cause Greig syndrome. H u m Mol Genet 1997; 6(11): 1979-1984. 2 2 5 . Zeller J, Schneider V, Malayaman S et al. Migration of zebrafish spinal motor nerves into the periphery requires multiple myotome-derived cues. Dev Biol 2002; 252(2):24l-256. 226. Karlstrom R O , Talbot W S , Schier AF. Comparative synteny cloning of zebrafish you-too: Mutations in the Hedgehog target gli2 affect ventral forebrain patterning. Genes Dev 1999; 13(4):388-393. 227. Hynes M, Stone D M , Dowd M et al. Control of cell pattern in the neural tube by the zinc finger transcription factor and oncogene Gli-1. Neuron 1997; 19(l):15-26. 228. Yang JT, Liu C Z , Villavicencio E H et al. Expression of human GLI in mice results in failure to thrive, early death, and patchy Hirschsprung-like gastrointestinal dilatation. Mol Med 1997; 3(12):826-835. 229. Ruiz i Altaba A. Combinatorial Gli gene function in floor plate and neuronal inductions by Sonic hedgehog. Development 1998; 125(12):2203-2212.

CHAPTER 9

Shh EKpression in Pulmonary Injury and Disease Paul M. Fitch, Sonia J. Wakelin, Jacqueline A. Lowrey, William A.H. Wallace and Sarah E.M. Howie* Abstract

T

he Hedgehog signalling pathway is crucial for normal vertebrate growth and development. Recent studies would suggest that signalling capability is retained in the post-embryonic organism. Shh signalling has been identified in the adult immune system, participating in CD4^ T lymphocyte activation. Studies on fibrotic pulmonary disorders have demonstrated Shh in both human and mouse lung restricted to areas of active disease. Acute lung injury has also shown upregulated expression and together this data is highly suggestive for a functional role for Shh signalling in adult lung injury and disease. We propose that hedgehog signalling may contribute to epithelial injury and repair and act as an intermediary in cross-talk between damaged epithelium and the immune/inflammatory system.

Introduction Work within our group has identified the up-regulated expression of the Hedgehog (Hh) signalling pathway in idiopathic pulmonary fibrosis (IPF) and murine models of fibrotic lung disease.^ That Hh signalling may also play a role in lung injury is also suggested by the work of Watkins et al '^ in their murine model of acute lung injury, and fiirther confirms a role for hedgehog signalling in post embryonic pulmonary tissues. Another area of interest to our group is the role of Hh signalling in the immune system. Active hedgehog signalling has been identified in cells of the peripheral immune system. ' " Since these immune cells are integral to inflammation and repair in the damaged lung, these observations have led to the emergence of the hedgehog signalling pathway as a possible target for therapeutic intervention in conditions of pulmonary damage and disease. The aim of this chapter is to review divergent avenues of research into the hedgehog signalling pathway and relate them to observations made with this pathway at the pulmonary interface.

The Hedgehog Pathway The hedgehog signalling pathway has been reviewed extensively elsewhere in this volume and in recent publications, thus will be covered only briefly in this review. There are three vertebrate hedgehog signalling molecules. Sonic hedgehog (Shh), Desert hedgehog (Dhh) and Indian hedgehog (Ihh). Of these, most studies have concentrated on the Shh signalling pathway. Shh mRNA generates an inactive 45kD precursor protein which auto-catalytically cleaves, *Corresponding Author: Sarah E.M. Howie—Immunobiology Group, MRC Centre for Inflammation Research, College of Medicine and Veterinary Medicine, University of Edinburgh, Edinburgh, U.K. Email: [email protected]


120


and is post-translationally modified to produce a highly hydrophobic cholesterol/palmitate modified Shh signalling molecule. This signals either as a cell surface molecule or is secreted following association with the Dispatched molecule. Binding of its receptor Patched (Ptc), releases Ptc mediated inhibition of Smoothened (Smo), a G protein coupled receptor-like molecule, allowing Smo to influence gene expression through the GLI zinc finger transcription factors. Signalling upregulates Ptc in an auto-regulatory loop, similar to many developmental pathways. Studies have shown Shh to be a crucial morphogen in a number of developmental systems, including the limbs, lung, gut, nervous and immune systems (for a review, see ref 8). It is only relatively recendy that a role for the hedgehog signalling pathway has been elucidated in the post embryonic organism.

Hedgehog Signalling in Normal Pulmonary Tissues Adult lung tissue contains several cell types, including Type I and II epithelial cells, Clara cells, mesenchymal cells (fibroblasts and endothelium), and resident immune system cells including alveolar and interstitial macrophages. Immunolocalisation studies from our group have demonstrated the expression of Shh protein in the lung is restricted to focal patches of epithelium in areas of injury and repair. Ptc is detectable on bronchial and alveolar epithelium and in alveolar macrophages.^ Watkins and colleagues have also identified normal expression of Ptc in the basal layer of bronchial epithelium.^ Continued expression of Ptc in the absence of ligand, might indicate regulation and/or signalling via another pathway.^ However, data emerging from studies in the gastrointestinal tract would suggest that Hh signalling may continue in tissue, in the apparent absence of immunodetectable ligand. Gastrointestinal tract Shh protein expression is localised to areas of regeneration, such as the fiindic glands of the stomach. '^^ Initially protein expression appeared to be absent from many areas of the adult mammalian tract such as the oesophagus, small intestine and colon, although some cells in these regions appeared to retain the ability to generate mRNA.^^'^^' Further to this, many of these areas lacked apparent Ptc expression. However, treatment of mice with cyclopamine, a Smo inhibitor, resulted in marked decreases in small intestine epithelial cell proliferation, as evidenced by BrdU incorporation and PCNA studies, suggesting active signalling." Of interest is the finding that in the mature colon, cyclopamine inhibition of Hh signalling results in a marked increase in colonic epithelial proliferation and reduced differentiation of the colonocytes. This alternate response to that of the small intestine may represent a Hh specific effect, with variation in Ihh/Shh expression, or may simply reflect the presence or absence of another signalling factor or downstream effector in the tissues of the colon. Explanations for signalling in the absence of ligand include the existence of another pathway upstream of Smo (indeed, Ptc-independent pathways have been postulated previously, ^^) or that Hh signalling continued, but at a concentration below immunohistochemical detection. Recent research would support the latter explanation, for Hh expression has now been immunolocalised to areas of the colon, previously described as negative. ^'^^ This was achieved through variations in antibody concentration and technique. Thus, future improvement in antibodies, detection methods and specific inhibitors may further clarify this issue for both the gastrointestinal and pulmonary systems.

Hedgehog Signalling in Injury and Disease Whilst the studies of Watkins et al demonstrated bronchiolar epithelial expression of the Shh signalling components, we have demonstrated Shh restricted to type II like cells in the alveoli of patients with IPR Expression was localised to areas of remodelling epithelia, particularly those areas with underlying fibrosis. Interestingly, whilst an irritant induced murine model of fibrosis illustrated similar localisation, a murine model of allergen induced inflammation

121

Shh Expression in Pulmonary Injury and Disease

A : SHH AS A WARNING SIGNAL Injurious Event

B : HYPOTHETICAL SYSTEM OF TYPE I CELL REPOPULATION

Mac

Epithelial Cells

Fibroblasts

Figure 1. A) Epithelial cell Shh has the potential to signal to neighboring cells to warn of cellular injury, and generate matrix deposition and proliferation in underlying fibroblasts, preventing transudation in the event of epithelial shedding. Communication with Macrophage and T lymphocytes is also possible, as is fibroblast-epithelial signaling. B) Type II cells are restricted to junctional positions at the entrance to alveoli, as shown this aerial diagram. In the event of type I cell loss, type II cells form intermediate cells which differentiate and spread into type I cells. Shh upregulation may represent a differentiation inhibitor, to which cells adjacent to denuded areas are nonresponsive. showed no such upregulation. Whilst both models illustrated evidence of inflammation, the allergen model lacked progressive fibrosis.^'^ T h e upregulation of Shh at sites of epithelial injury comes amongst a plethora of other modulating signals and could therefore represent a downstream marker of a m u c h larger process initiated by the underlying mesenchymal/fibroblastic cells, which play a crucial role in epithelial maintenance. Mesenchymal cells produce specific epithelial growth factors such as keratinocyte growth factor (KGF) a n d fibroblast growth factor 10 ( F G F - 1 0 ) . These are upregulated in response to injurious stimuli, resulting in increased epithelial proliferation and rapid reestablishment of an intact epithelial layer (for a review, see ref. 17). Studies using FGF-10, F G F R 2 b and Shh knockout mice have recently confirmed Shh to be downstream of F G F - 1 0 signalling. ^^ Shh expression on epithelium at sites of fibroblastic proliferation, such as that observed in IPF, may be a consequence of this repair process rather than an instigator of repair. T h u s , Shh could be considered a regulatory factor in this instance, given its ability to downregulate FGF-10^^ (see Fig. l A ) . Alternatively expression at the alveolar surfaces of IPF patients might relate to a specific element of the disease process, which maybe mirrored in the F I T C mouse model, where we observe a similar pattern of expression.^ IPF is a chronic fibrotic condition of the lung, associated with inflammation, where median survival after diagnosis is limited to five years. Although without k n o w n cause, progression is believed to have an immunological basis, linked in some way with aberrant epithelial mesenchymal interaction.^^ Previous work in this laboratory identified circulating antibodies to a 7 0 - 9 0 K D protein in IPF patient serum, which were n o t observed in normal controls. This protein was later localised to type II cells (Fig. 2), suggesting an autoantigen. ^ Experiments in vivo with a polyclonal antibody raised against the antigen and a h u m a n type II epithelial cell line showed upregulated tenascin a n d T G F - P production. Tenascin is an extracellular matrix protein associated w i t h active scar formation, whilst T G F - p can be pro-fibrotic. Both have been well characterised as being present in the lungs of patients with j p p 2 ,25 ^ ^ j tenascin has been shown to be localised to areas of active fibroplasias or 'fibroblastic foci'. It will be interesting to discover whether Shh is also upregulated by this interaction, as it would explain its localised expression, at areas of putative antibody interaction.


122

BRONCHI & BRONCHIOLES

ALVEOLI

SMALL AND TERMINAL BRONCHIOLES

Type I Epithelium

"^^^^^^^

- 95% of Surface Area - 4 0 % of Cell No - Gas Exchange Surface NeuroEpithella! Body

Stratified Epithelium Mucous Glands f ^ & Cells - Progenitor Function Ciliated & Goblet cells Basal Cells

(§

- Progenitor Function NeuroEpithelial Body - Progenitor Function

~ Vulnerable to Injury

iMnacffloffl Clara Cells

|

- Secretory Function

Type II Epithelium

Variant Clara Cells

- 5% of Surface Area

i

0

- Progenitor Function

- 6 0 % of Cell No

Neuroendocrine Cells y

- Progenitor & Secretory Function

~ Progenitor Function

No Neuroendocrine Cells

No Mucus Cells/Glands

No Mucous Cells/Glands

Figure 2. Cell types at the pulmonary surface. Shh also has the potential to be profibrotic, Shh and FGF signalling are well linked in neural systems both pre and postnatally^^' and hedgehogs have been linked with collagen fibre induction in the neural system. Shh also has substantial cell cycle modulating capabilities^^'^^ and participates in well characterised proliferative mesenchymal interactions in the developing lung. Thus the localisation of Shh at fibrotic foci in IPF may also present a causal relationship with fibrosis in the IPF patient. Similar hypotheses can be drawn from the observations made in the fibrotic FITC mouse model of IFF pathology. Fibrosis and inflammation are believed to be induced in this model through the persistence of the intratracheally instilled irritant FITC, which lodges in the interstitium and binds resident proteins. Antibodies are raised to FITC, and mice develop a pathology similar to that of IPF patients, where Shh expression is localised to sites of damage. Interestingly, recent work has illustrated a requirement for cell confluence for Shh responsiveness, in a Ptc expressing Shh bioassay.^^'^^ Thus, Shh and Ptc expression at a site of proliferative repair, may not automatically indicate that active signalling is occurring. Indeed, the broad expression area of Ptc in the lung might represent a "Shh sponge", to endocytose, in the absence of signalling, free Shh, as described by Torroja and colleagues.^ However, whilst this may, in part, be true for some expression studies in the lung, work performed in our laboratory and that of others has identified alterations in target gene expression such as Ptc and Glil, coincident with Shh expression, suggesting that at least some active signalling results from Shh expression.^ Shh would appear to share many of its effects with KGF and FGF-10,^^ most notably, stimulation of increased cellular resistance to injury and death.^^'^ Shh exhibits this effect in a number of cell types,^^'^^'^^ but has yet to be established at the pulmonary interface. Perhaps the best example of this activity is in Parkinson's disease, where exogenous administration of Shh in both a marmoset and mouse model of the disease results in improved locomotion and function. ' Tsuboi and colleagues, citing work by Miao et al, have suggested that this is


123

likely due to Shh acting as a neuroprotective agent against dopaminogenic neuron cell death, which is central to disease progression. In support of this Thilbert et al have shown that Shh can induce the anti-apoptotic protein Bcl-2 in keratinocytes, and that its presence prevents the apoptosis of neuroepithelial cells. Therefore in a pulmonary situation of injury, such as exposure to environment oxidants, Shh release, along with KGF and FGF-10, may serve to temporarily increase surrounding epithelial cell resilience, and prevent complete denudation of an epithelial surface and the risk of secondary infection (Fig. lA). Sustaining viability of injured cells carries with it the increased probability of populating the epithelium with cells which have incurred genetic damage. Indeed, this may help to explain the higher incidence of small and nonsmall cell lung cancers following epithelial trauma, and the preponderance for the involvement of Shh signalling in these carcinomas. Upregulation of Hh signalling following injury, can be found in a number of model systems Pola et al "^^ ^ have demonstrated the upregulation of Shh and Ptcl expression in a rodent hind limb ischemia model. Furthermore, the same group demonstrated that the administration of Shh induced robust neovascularisation, enhanced blood flow, and the upregulation of the angiogenic factors, vascular endothelial growth factor (VEGF) and the angiopoietins, Angl and Ang2, ^ although any direct association is a matter of conjecture. ' ^ The findings from other groups have been broadly in keeping with those of Pola et al demonstrating that Shh is capable of inducing VEGF expression and angiogenesis. Such angiogenic modulation could have important implications at the respiratory surfaces of IPF patients, where we have observed Shh expression in association with fibrotic foci. Shh expression here may represent an attempt by the lung to revascularise and remodel the fibrotic architecture. Alternatively, it may represent a step in the progression of the fibrosis, where VEGF itself is found upregulated with epithelial injury. Perhaps the greatest implication is in the neoangiogenic advancement of pulmonary malignancies, which are associated with the expression of Hh pathway components.

Shh in the Immune System A functional role for hedgehog signalling in the adult immune system is perhaps not surprising given the well documented role of Shh in lymphoid development and differentiation. ' ' Work in our laboratory has found evidence for expression of Shh and Ptc in both resting and activated murine and human CD4^ and CD8^ T lymphocytes, human macrophages and secondary lymphoid tissue, ' both at the mRNA and protein level. This lends support to a possible role for immune-epithelial cell interaction at sites of injury and repair. It has been postulated that the expression of Hh in a large range of diversified tissues and cell types, might represent a common method of sensing the cellular environment and neighbouring cell health. Given the expression of the Ptc molecule on human macrophages, it may be postulated that Shh expression acts as an advertisement of epithelial damage both to the macrophage, to facilitate attention, and to the adjacent cells as a warning signal for defence and repair initiation (Fig. lA). Further to this, Lowrey and coworkers^' have demonstrated that exogenous Shh has the capacity to push sub-optimally activated peripheral CD4^ T lymphocytes through further cell cycles, whilst addition of a blocking antibody to endogenous Shh results in decreased proliferation. Together these results would suggest that exogenous Shh administration enhances an endogenous stimulatory effect of Shh. Indeed, further work from our group has demonstrated that Shh can act as a costimulatory molecule to T-cells activated via the T cell receptor using anti-CD3 antibodies (Lowrey et al, unpublished observations). This explains the upregulation of the activation antigens CD25 and CD69, by exogenous Shh, along with the production of Interleukin 2 (IL-2), a T lymphocyte proliferative factor, interleukin 10 (IL-10) and Interferon gamma (IFN-y). Thus Shh expression at sites of disease and damage has the potential to sustain

124


T lymphoqê mediated immune responses and even to influence the effector function of immune responses, through an IFN-y bias. Shh has also been shown to upregulate Bcl-2 in T lymphocytes, a step linked with memory T lymphocyte generation, and thus, Shh at sites of injury might also serve to ensure adaptive immunity to damaging antigens, or an exacerbation of autoimmunity. Given that many of the diseases associated with Shh expression contain elements of autoimmunity or immune dysregulation, this crosstalk could have important implications both for the initiation and maintenance of a disease process.

Hedgehogs and Pulmonary Progenitor Cells Watkins and colleagues^ depleted Clara cells from mouse airways via their sensitivity to naphthalene injury. This resulted in a transient peak in Shh and Gli expression 3 days post depletion that coincided with the reestablishment of a complete epithelial barrier. It was suggested that Shh signalling might be involved in this repopulation step, since this peak preceded an increase in the number of pulmonary neuroepithelial cells—a cell type identified as a potential epithelial progenitor. ^'^ Subsequendy, Watkins and colleagues identified clusters of epithelial cells expressing Ptc and the neuroendocrine marker. Calcitonin Gene Related Peptide (CGRP), associated with Shh expressing cells in the developing lung.^ It was suggested that this clustered interaction might continue into the adult pulmonary system, where Shh expression might influence neuroepithelial differentiation. Presendy, progenitor cell phenotypes are characterised via their longevity via nucleotide analogue incorporation, and are defined as label retaining cells. In the mouse these cells are localised to glandular submucosal glands in the distal trachea and bronchi.^^ In the absence of these glands, such as in the bronchioles and alveoli, LRCs occur in localised foci,^^ referred to as neuroepithelial bodies (Fig. 2). These foci are populated by Clara Cell secretory protein (CCSP)-expressing epithelial cells, known as variant Clara cells (vCC), which are naphthalene resistant. Associated with these are CGRP-expressing epithelial cells which show intermediate morphology between a pulmonary Neuroendocrine Cells (PNEC) and a Clara cell '^^ (Fig. 2). Infiltration of blood borne progenitors in this regenerative process, add a further level of complexity outwith the scope of this chapter, but covered elsewhere. ^^ Label retaining cells undergo hyperplasia in the event of Clara cell death, such as that induced by naphthalene, resulting in the rapid reestablishment of a functional epithelial cell layer. Selective depletion of vCC using a Herpes simplex virus thymidine kinase linkage system, prior to naphthalene injury, results in a failure to repopulate a differentiated ciliated epithelium, despite PNEC hyperplasia. Whether this highlights the CCSP expressing cells as progenitor cells, producers of a signal necessary for PNEC progenitor cell function or as coprogenitors with the PNEC is still an area of conjecture. As to a potential role for Shh in this process, epithelial coverage is completed at the peak of Shh expression, thus a causative role in nonspecific epithelial proliferation is unlikely, and this is confirmed in our studies where we observe bronchiolar Shh expression restricted to areas of complete epithelial coverage. It is interesting that coverage normally occurs in the absence of differentiation, which typically occurs following reestablishment of epithelial integrity. Given the necessity for confluence for Shh signalling in some cell lines, as mentioned previously, one might speculate that it is in this differentiation step that Shh might play a role. Reasoning for such speculation arises through the well characterised progenitor modulating role of Hh in a range of post embryonic systems including T lymphocyte development and haematopoiesis. The work by Bhardwaj in haematopoiesis was notable for it clearly defined the downstream effector of Shh function as BMP-4, where Shh addition to an enriched population of human CD34+Lin-CD38- progenitor cells resulted in increased self renewal, albeit in combination with many other growth factors. An example of Hh mediated progenitor modulation in injury comes indirectly from several sources. ' Pepinsky et al demonstrated that Shh administration can accelerate nerve recovery


125

following sciatic nerve crush injury.^^ A Possible explanation for this response comes from studies by Bambakidos et al, using a rat demyelination model. These authors observed increased proliferation of stem cell like progenitors in areas of chemical demyelination following direct Shh administration. This suggests that exogenous Shh at a site of injury induces proliferation in stem cell populations, although the specificity of this progenitor proliferation was not addressed in this study. The manner in which hedgehog signalling might facilitate this stem cell influence has yet to be characterised, but is likely to lie in its ability to regulate differentiation. Evidence in support of such a role comes from studies of fracture repair, in which Ihh may play a role. Following a fracture, pluripotent mesenchymal cells invade and differentiate into osteoblasts, to initiate the production of hard callus comprising bone matrix, and chondrocytes for the formation of the soft callus. A balanced secondary chondrocyte differentiation step facilitates the remodelling of bone until the bone shape is restored. '^^ Ihh expression is induced by 3 days post fracture persisting past 2 weeks in a murine model, with the greatest level of expression found in chondrocytes undergoing their secondary differentiation step.^^' Ihh expression here induces further chrondrocyte proliferation, but prevents further chondrocyte differentiation via PTHrP signalling. " Given the osteoblast upregulation of PTHrP in the presence of rShh shown by Jemdand and colleagues and the osteoblastic lineage bias induced by Ihh,^ it is likely that in this system, Ihh is central to restricting differential fates of invading pluiripotent mesenchymal cells and facilitates an accurate bone remodelling response. Central to the stem cell maintenance function of Shh observed in haematopoiesis, is its ability to upregulate BMP-4. These molecules are crucial in early lung morphogenesis and thus have substantial signalling potential in the post embryonic lung. Indeed, studies in a mouse model of allergic inflammation have demonstrated an upregulation of BMP s at sites of inflammation, including BMP-4.^^ In vitro studies would suggest that BMP-4 has anti-proliferative effects in cancer derived cell lines, however these effects can often be contradictory and dependant on the cosignals and cell types present in culture. In vivo studies into pulmonary BMP functions are limited by the lethality of BMP-4 knockouts. However, innovative in vivo studies in this field from the lab of Brigid Hogan and colleagues have demonstrated a potential progenitor modulating function for BMP*s in epithelium. Cells exposed to high levels of BMP-4 retained undifferentiated characteristics. ' In light of these studies, were Shh to modulate progenitor function through the BMP's, it would likely be in conjunction with a number of other signals, such as FGF-10. Perhaps in a recapitulation of the process of cell fate designation in the embryonic lung.

Shh and Type II Epithelial Cells Studies thus far have focused on roles for Shh expression at the bronchi and bronchiolar levels, however immunohistochemical data suggest that the hedgehog signal might persist in the larger and smaller airway systems. Watkins and colleagues observed a persistence of Ptc signalling in the progenitor cells of the trachea and bronchi, whilst our lab has identified expression of Shh and Ptc in type II like cells of the terminal bronchioles and alveoli (Fig. 2). Pulmonary epithelial regeneration occurs rapidly at the alveolar level. Here, in response to the removal of contact inhibition and KGF,^ type II cells downregulate specific functional machinery, proliferate and spread out to become thinly spread type I cells, specialised for gaseous exchange. This process occurs almost continually as type I cells are highly susceptible to injury by exogenous factors, i.e., pneumotoxic environmental pollutants and by factors involved in both innate and adaptive pathogen clearance in the lung. Whilst lack of contact inhibition and KGF are triggers for type II differentiation, no maintenance signal for type II cell number has yet been identified. Perhaps Shh has a modulating function here too. Certainly immunolocalisation in IPF patient biopsies has identified Shh upregulation in individual type II like cells amongst adjacent negative cells, with expression isolated to areas of injury.^

126


This makes for an attractive hypothesis. Damage upregnlates type II cell expression of Shh, this induces type II proliferation and inhibits differentiation. However, type II cells adjacent to areas of denuded basement membrane have incomplete confluence and thus lose responsiveness to the Shh, allowing them to differentiate and cover the exposed area (Fig. IB). However, were Shh a maintenance factor in this continual type II-I transition, it might be expected that normal lung might exhibit some limited immunohistochemical reactivity, and this is not observed. Whether this represents true absence of Shh protein, or the limits of immunohistochemical detection, remains to be determined.

Concluding Comments The potential for post embryonic recapitulation of developmental signals in injury and disease presents many exciting targets for therapeutic modulation. This is particularly true for the pulmonary system, where the air interface facilitates the direct and rapid delivery of short half life Hh modulating agents to target cells. This will avoid the systemic complications of treatment that might be expected in the treatment of conditions such as multiple sclerosis and Parkinson's disease. Our knowledge of the Hh signalling pathway and its role in development has made great advancements over recent years and has led to the development of many new and exciting ideas which may identify functions for the Hh pathway in post-embryonic systems.

References 1. Stewart GA, Hoyne GF, Ahmad SA et al. Expression of the developmental Sonic hedgehog (Shh) signalling pathway is up-regulated in chronic lung fibrosis and the Shh receptor patched 1 is present in circulating T lymphocytes. J Pathol 2003; 199(4):488-495. 2. Watkins DN, Berman DM, Baylin SB. Hedgehog signaling: progenitor phenotype in small-cell lung cancer. Cell Cycle 2003; 2(3):196-198. 3. Watkins DN, Berman DM, Burkholder SG et al. Hedgehog signalling within airway epithelial progenitors and in small-cell lung cancer. Nature 2003; 422(6929):313-317. 4. Benson RA, Lowrey JA, Lamb JR et al. The Notch and Sonic hedgehog signalling pathways in immunity. Mol Immunol 2004; 41 (6-7):715-725. 5. Lowrey JA, Stewart GA, Lindey S et al. Sonic hedgehog promotes cell cycle progression in activated peripheral CD4(+) T lymphocytes. J Immunol 2002; 169(4):1869-1875. 6. Stewart GA, Lowrey JA, Wakelin SJ et al. Sonic hedgehog signaling modulates activation of and cytokine production by human peripheral CD4+ T cells. J Immunol 2002; 169(10):5451-5457. 7. Nybakken K, Perrimon N. Hedgehog signal transduction: Recent findings. Curr Opin Genet Dev 2002; 12(5):503-511. 8. McMahon AP, Ingham PW, Tabin CJ. Developmental roles and clinical significance of hedgehog signaling. Curr Top Dev Biol 2003; 53:1-114. 9. Agren M, Kogerman P, Kleman MI et al. Expression of the PTCHl tumor suppressor gene is regulated by alternative promoters and a single functional Gli-binding site. Gene 2004; 330:101-114. 10. van den Brink GR, Hardwick JC, Nielsen C et al. Sonic hedgehog expression correlates with fundie gland differentiation in the adult gastrointestinal tract. Gut 2002; 51(5):628-633. 11. van den Brink GR, Hardwick JC, Tytgat GN et al. Sonic hedgehog regulates gastric gland morphogenesis in man and mouse. Gastroenterology 2001; 121(2):317-328. 12. Dimmler A, Brabletz T, Hlubek F et al. Transcription of sonic hedgehog, a potential factor for gastric morphogenesis and gastric mucosa maintenance, is up-regulated in acidic conditions. Lab Invest 2003; 83(12):1829-1837. 13. Ishizuya-Oka A, Ueda S, Inokuchi T et al. Thyroid hormone-induced expression of sonic hedgehog correlates with adult epithelial development during remodeling of the Xenopus stomach and intestine. Differentiation 2001; 69(l):27-37. 14. van den Brink GR, Bleuming SA, Hardwick JC et al. Indian Hedgehog is an antagonist of Wnt signaling in colonic epithelial cell differentiation. Nat Genet 2004; 36(3):277-282. 15. Testaz S, Jarov A, Williams KP et al. Sonic hedgehog restricts adhesion and migration of neural crest cells independently of the Patched- Smoothened-Gli signaling pathway. Proc Natl Acad Sci USA 2001; 98(22):12521-12526. 16. Roberts SN, Howie SE, Wallace WA et al. A novel model for human interstitial lung disease: Hapten-driven lung fibrosis in rodents. J Pathol 1995; 176(3):309-318.

Shh Expression in Pulmonary

Injury and Disease

127

17. Ware LB, Matthay MA. Keratinocyte and hepatocyte growth factors in the lung: Roles in lung development, inflammation, and repair. A m J Physiol Lung Cell Mol Physiol 2002; 282(5) :L924L940. 18. Rice R, S p e n c e r - D e n e B, C o n n o r E C et al. D i s r u p t i o n of F g f l 0 / F g f r 2 b - c o o r d i n a t e d epithelial-mesenchymal interactions causes cleft palate. J Clin Invest 2004; 113(12):1692-1700. 19. Chuang P T , Kawcak T , M c M a h o n AP. Feedback control of mammalian Fledgehog signaling by the Hedgehog-binding protein. H i p 1, modulates Fgf signaling during branching morphogenesis of the lung. Genes Dev 2003; 17(3):342-347. 20. Thannickal VJ, Toews GB, White ES et al. Mechanisms of pulmonary fibrosis. Annu Rev M e d 2004; 55:395-417. 2 1 . Wallace WA, Roberts SN, Caldwell H et al. Circulating antibodies to lung protein(s) in patients with cryptogenic fibrosing alveoHtis. Thorax 1994; 49(3):218-224. 22. Wallace WA, Schofield JA, Lamb D et al. Localisation of a pulmonary autoantigen in cryptogenic fibrosing alveolitis. Thorax 1994; 49(11):1139-1145. 2 3 . Wallace WA, Howie SE. Upregulation of tenascin and T G F b e t a production in a type II alveolar epithelial cell Hne by antibody against a pulmonary auto-antigen. J Pathol 2 0 0 1 ; 195(2):251-256. 24. Wallace WA, Howie SE, Lamb D et al. Tenascin immunoreactivity in cryptogenic fibrosing alveolitis. J Pathol 1995; 175(4):4l5-420. 25. Khalil N , O ' C o n n o r RN, U n r u h H W et al. Increased production and immunohistochemical localization of transforming growth factor-beta in idiopathic pulmonary fibrosis. A m J Respir Cell Mol Biol 1991; 5(2):155-162. 26. Wechsler-Reya RJ, Scott M P . Control of neuronal precursor proliferation in the cerebellum by Sonic Hedgehog. Neuron 1999; 22(1):103-114. 27. Ruiz i Altaba A, Palma V, Dahmane N . Hedgehog-Gli signalling and the growth of the brain. N a t Rev Neurosci 2002; 3 ( l ) : 2 4 - 3 3 . 28. Brewster R, MuUor JL, Ruiz i Altaba A. Gli2 functions in FGF signaling during antero-posterior patterning. Development 2000; 127(20):4395-4405. 29. Parmantier E, Lynn B, Lawson D et al. Schwann cell-derived Desert hedgehog controls the development of peripheral nerve sheaths. Neuron 1999; 23(4):713-724. 30. Fan H , Khavari PA. Sonic hedgehog opposes epithelial cell cycle arrest. J Cell Biol 1999; 147(1):717G. 3 1 . Sherr CJ, Roberts J M . C D K inhibitors: positive and negative regulators of Gl-phase progression. Genes Dev 1999; 13(12):1501-1512. 32. Ingram W J , Wicking CA, G r i m m o n d SM et al. Novel genes regulated by Sonic Hedgehog in pluripotent mesenchymal cells. Oncogene 2002; 21(53):8196-8205. 33. Taipale J, Chen JK, Cooper M K et al. Effects of oncogenic mutations in Smoothened and Patched can be reversed by cyclopamine. Nature 2000; 406(6799): 1005-1009. 34. Torroja C, Gorfinkiel N , Guerrero I. Patched controls the Hedgehog gradient by endocytosis in a dynamin-dependent manner, but this internalization does not play a major role in signal transduction. Development 2004; 131(10):2395-2408. 35. U p a d h y a y D , B u n d e s m a n n M , P a n d u r i V et al. Fibroblast g r o w t h factor-10 a t t e n u a t e s H 2 0 2 - i n d u c e d alveolar epithelial cell D N A damage: Role of MAPK activation and D N A repair. Am J Respir Cell Mol Biol 2004; 31(1):107-113. 36. Oswari J, Matthay MA, Margulies SS. Keratinocyte growth factor reduces alveolar epithelial susceptibility to in vitro mechanical deformation. A m J Physiol Lung Cell Mol Physiol 2 0 0 1 ; 281(5):L1068-L1077. 37. Thibert C, Teillet MA, Lapointe F et al. Inhibition of neuroepithelial patched-induced apoptosis by sonic hedgehog. Science 2003; 301(5634):843-846. 38. Miao N , W a n g M , O t t JA et al. Sonic hedgehog promotes the survival of specific C N S neuron populations and protects these cells from toxic insult In vitro. J Neurosci 1997; 17(15):5891-5899. 39. Dass B, Iravani M M , Jackson MJ et al. Behavioural and immunohistochemical changes following supranigral administration of sonic hedgehog in l-methyl-4-phenyl-1,2,3,6-tetrahydropyridine-treated common marmosets. Neuroscience 2002; 114(1):99-109. 40. Tsuboi K, Shults C W . Intrastriatal injection of sonic hedgehog reduces behavioral impairment in a rat model of Parkinson's disease. Exp Neurol 2002; 173(1):95-104. 4 1 . Reynolds SD, Giangreco A, Power J H et al. Neuroepithelial bodies of pulmonary airways serve as a reservoir of progenitor cells capable of epithelial regeneration. A m J Pathol 2000; 156(l):269-278. 42. Pola R, Ling LE, Aprahamian T R et al. Postnatal recapitulation of embryonic hedgehog pathway in response to skeletal muscle ischemia. Circulation 2 0 0 3 ; 108(4):479-485. 43. Pola R, Ling LE, Silver M et al. T h e morphogen Sonic hedgehog is an indirect angiogenic agent upregulating two families of angiogenic growth factors. N a t Med 2 0 0 1 ; 7(6):706-711.

128

Hedgehog-Gli

Signaling in Human

Disease

44. Lawson N D , Vogel A M , Weinstein BM. Sonic hedgehog and vascular endotheUal growth factor act upstream of the N o t c h pathway during arterial endothelial differentiation. Dev Cell 2002; 3(1):127-136. 4 5 . Kanda S, Mochizuki Y, Suematsu T et al. Sonic hedgehog induces capillary morphogenesis by endothelial cells through phosphoinositide 3-kinase. J Biol Chem 2003; 278(10):8244-8249. 46. Pham I, Uchida T, Planes C et al. Hypoxia upregulates VEGF expression in alveolar epithelial cells in vitro and in vivo. A m J Physiol Lung Cell Mol Physiol 2002; 283(5):L1133-L1142. ^1. O u t r a m SV, Varas A, Pepicelli C V et al. Hedgehog signaling regulates differentiation from double-negative to double-positive thymocyte. Immunity 2000; 13 (2): 187-197. 48. Varas A, Hager-Theodorides AL, Sacedon R et al. T h e role of morphogens in T-cell development. Trends Immunol 2003; 24(4): 197-206. 49. Ruiz i Altaba A, Stecca B, Sanchez P. Hedgehog—Gli signaling in brain tumors: Stem cells and paradevelopmental programs in cancer. Cancer Lett 2004; 204(2): 145-157. 50. Reynolds S D , H o n g KU, Giangreco A et al. Conditional clara cell ablation reveals a self-renewing progenitor function of pulmonary neuroendocrine cells. Am J Physiol Lung Cell Mol Physiol 2000; 278(6):L1256-L1263. 5 1 . Borthwick D W , Shahbazian M, Krantz Q T et al. Evidence for stem-cell niches in the tracheal epithehum. Am J Respir Cell Mol Biol 2 0 0 1 ; 24(6):662-670. 52. H o n g KU, Reynolds SD, Giangreco A et al. Clara cell secretory protein-expressing cells of the airway neuroepithelial body microenvironment include a label-retaining subset and are critical for epithelial renewal after progenitor cell depletion. Am J Respir Cell Mol Biol 2 0 0 1 ; 24(6):671-681. 53. Giangreco A, Shen H , Reynolds SD et al. Molecular phenotype of airway side population cells. A m J Physiol Lung Cell Mol Physiol 2004; 286(4):L624-L630. 54. Bhardwaj G, Murdoch B, W u D et al. Sonic hedgehog induces the proliferation of primitive human hematopoietic cells via BMP regulation. N a t Immunol 2 0 0 1 ; 2(2):172-180. 55. Pepinsky RB, Shapiro RI, Wang S et al. Long-acting forms of Sonic hedgehog with improved pharmacokinetic and pharmacodynamic properties are efficacious in a nerve injury model. J Pharm Sci 2002; 91(2):371-387. 56. Bambakidis N C , W a n g RZ, Franic L et al. Sonic hedgehog-induced neural precursor proliferation after adult rodent spinal cord injury. J Neurosurg 2003; 99(1 Suppl):70-75. 57. Bolander M E . Regulation of fracture repair by growth factors. Proc Soc Exp Biol Med 1992; 200(2):165-170. 58. Sandberg M M , Aro H T , Vuorio EI. Gene expression during bone repair. Clin Orthop 1993; (289):292-312. 59. Ferguson C M , Miclau T, H u D et al. C o m m o n molecular pathways in skeletal morphogenesis and repair. Ann NY Acad Sci 1998; 857:33-42. 60. Vortkamp A, Pathi S, Peretti G M et al. Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech Dev 1998; 71(l-2):65-76. 6 1 . St Jacques B, Hammerschmidt M, McMahon AP. Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev 1999; 13(16):2072-2086. 62. Lanske B, Karaplis A C , Lee K et al. P T H / P T H r P receptor in early development and Indian hedgehog-regulated bone growth. Science 1996; 273(5275):663-666. 6 3 . Zou H , Wieser R, Massague J et al. Distinct roles of type I bone morphogenetic protein receptors in the formation and differentiation of cartilage. Genes Dev 1997; ll(17):2191-2203. 64. Jemtland R, Divieti P, Lee K et al. Hedgehog promotes primary osteoblast differentiation and increases P T H r P m R N A expression and i P T H r P secretion. Bone 2003; 32(6):611-620. 65. Spinella-Jaegle S, Rawadi G, Kawai S et al. Sonic hedgehog increases the commitment of pluripotent mesenchymal cells into the osteoblastic lineage and abolishes adipocytic differentiation. J Cell Sci 2 0 0 1 ; l l 4 ( P t l l ) : 2 0 8 5 - 2 0 9 4 . Gd. Rosendahl A, Pardali E, Speletas M et al. Activation of bone morphogenetic protein/Smad signaling in bronchial epithelial cells during airway inflammation. A m J Respir Cell Mol Biol 2002; 27(2): 160-169. GJ. Buckley S, Shi W , Driscoll B et al. BMP4 signaling induces senescence and modulates the oncogenic phenotype of A549 lung adenocarcinoma cells. Am J Physiol Lung Cell Mol Physiol 2004; 286(1):L81-L86. 68. Weaver M, Yingling J M , D u n n N R et al. Bmp signaling regulates proximal-distal differentiation of endoderm in mouse lung development. Development 1999; 126(18):4005-4015. 69. Bellusci S, Henderson R, Winnier G et al. Evidence from normal expression and targeted misexpression that bone morphogenetic protein (Bmp-4) plays a role in mouse embryonic lung morphogenesis. Development 1996; 122(6): 1693-1702.

CHAPTER 10

Human G)rrelates of GLI3 Function Leslie G. Biesecker* Abstract

T

he human phenotypes caused by mutations in GLI3 are a prototype for the utility of clinical analysis to unravel gene function. When the human studies were coupled with basic research, the pathway to understanding gene function was easier to discern and more rapidly accomplished. Oudining the series of discoveries that led to our current understanding of GLI3 and genotype-phenotype correlation is useful to demonstrate how these discoveries can be facilitated in the future.

The Delineation of the Greig Cephalopolysyndactyly Syndrome (GCPS) The story of GLI3 and human pathology began with a 1926 publication^ by the Edinburgh surgeon David Middleton Greig.* He described a series of patients with a clinical finding of "oxycephaly", defining this term as "general craniofacial synostosis". He recognized that these craniofacial findings often occurred with limb anomalies. However, he lumped a number of patients together who probably had what we now recognize to be Carpenter, Crouzon and Saethre-Chotzen syndromes into the category of "oxycephaly". One case (case VI) in his report also had cranial and limb anomalies but was described as an exception to the general concept of "oxycephaly". This girl had a "broad and voluminous forehead", her hands had short, broad thumbs with terminal phalangeal lacunae and numerous scars from previous attempts to correct cutaneous syndactyly, and her feet demonstrated broad great toes with scars from removal of a partially duplicated distal phalanges, with complete cutaneous syndactyly of toes 1 to 3 and partial syndactyly of toes 3 and 4. She was "considered to be rather above the average at school". Her mother was said to have "symmetrical webbing of the fingers of both hands" but no craniofacial findings. Greig mistakenly cleaved the craniofacial findings from the limb findings in this case: "The webbing of the digits was hereditary and the head affection a dysplasia of the basis cranii." He did not apply the concept of pleiotropy nor that of variable expressivity to his analysis. The application of current concepts of syndrome delineation lead us to conclude that this patient and her mother were affected by the disorder we now consider to be GCPS. In spite of the limitations of Greig's observations and analysis, the designation of "Greig cephalopolysyndactyly syndrome" was coined to honor this description. Ironically, Greig rejected the use of the compound term "acrobrachycephalosyndactylism", a term that is similar to that which we now combine with his name. * The surname is pronounced "greg", with a trilled r. The surname of the composer Edward Grieg is pronounced "greeg". *Leslie G. Biesecker—National Human Genome Research Institute, NIH, Building 49, Rm. 4A80, Bethesda, Maryland 20892 U.S.A. Email: [email protected]

Hedgehog-Gli Stealing in Human Disease^ edited by Ariel Ruiz i Altaba. ©2006 Eurekah.com and Springer Science+Business Media.

130


The heritable nature of the condition was affirmed by a number of authors, including a family often patients in four generations."^ Primarily through the approach of clinical analysis of affected patients, it was confirmed that the typical manifestations included hypertelorism,* broad or prominent forehead, macrocephaly, with polysyndactyly. The Polydactyly was most commonly preaxial Polydactyly of the feet and postaxial Polydactyly of the hands,** and varying degrees of cutaneous syndactyly, most commonly of digits 3 and 4 in the hands and when present, often involves all five digits of the feet. '^ Other less common manifestations include craniosynostosis (once thought to be a common manifestation), mild mental retardation, umbilical and inguinal hernias, hydrocephalus, and agenesis or hypoplasia of the corpus callosum.

The Molecular Basis of GCPS The first insight into the molecular basis of GCPS was the observation that patients with interstitial deletions of the short arm of chromosome 7 manifested the syndrome. '^ These cytogenetic localizations were refined by molecular techniques and led to the determination that GCPS was caused by interruption of the GLI3 zinc finger transcription factor gene,^ presumably through the mechanism of haploinsufficiency. The study of humans with GLI3 mutations lay relatively dormant from that time up to the discovery of the etiology of Pallister-Hall syndrome.

The Clinical Delineation of Pallister-Hall Syndrome (PHS) We set out to rigorously delineate the clinical features of PHS in preparation to undertake a positional cloning effort. The PHS phenotype was originally described in a cohort of six individuals who had Polydactyly (typically central Polydactyly^ and sometimes postaxial Polydactyly) and hypothalamic hamartoma with varying degrees of pituitary dysfunction, a shortened nose, imperforate anus, pulmonary segmentation anomalies, and laryngotracheal anomalies ranging from bifid epiglottis to laryngotracheal clefts. ^^ In the early 1990s several case reports demonstrated that PHS was also inherited in an autosomal dominant pattern.^ ^' Subsequendy, clinical diagnostic criteria were developed that specified that a proband had to have central Polydactyly and a hypothalamic hamartoma. ^^ Using this strict clinical definition, a positional cloning effort was undertaken on several large families with PHS, which led to its mapping to 7pl4, in a small candidate region that included GLI3. After all other genes were excluded, GLI3 was sequenced, and frameshift mutations were found in two families. ^^ At this point, the question arose as to whether PHS and GCPS were distinct clinical entities. To approach this question, it is necessary to review some basic principles of clinical dysmorphology. First, pleiotropic developmental anomaly syndromes can be identified that are distinct clinical entities. Second, all syndromes have anomalies that are shared with some other syndromes, that is, they have overlapping manifestations. To define a distinct pleiotropic anomaly syndrome, one has to identify a pattern of anomalies that is typical for the disorder and demonstrate that it is distinct from existing syndromes. The anomalies described above seemed to cleave PHS and GCPS into two distinct syndromes (Table 1). From these data, it appeared that GCPS and PHS had little overlap, although preaxial Polydactyly was found in occasional patients with PHS, and postaxial Polydactyly was * Hypertelorism is an abnormally large interpupillary distance, commonly confused with telecanthus, which is an abnormally large distance between the inner canthi. Curiously, Greig wrote a paper on hypertelorism in 1924, but did not mention this finding in the "oxycephaly" case description in 1926. ** Duplication of the thumb or great toe in humans is called preaxial Polydactyly, while in developmental biology it is called anterior Polydactyly. * Central Polydactyly is synonymous with insertional or mesoaxial Polydactyly. % A hamartoma is a focal malformation that is composed of an abnormal mixture of tissue elements. It is not a neoplasm.

Human Correlates ofGLId Function

131

Table 1. Distinct clinical manifestations in GCPS and PHS

Hypertelorisim Preaxial Polydactyly Postaxial Polydactyly Hypothalamic hamartoma Central Polydactyly Imperforate anus

GCPS

PHS

+++ +++ ++ -

+ ++ +++ +++ +

+++ common manifestation; ++ seen in some cases; + seen uncommonly; - not reported

common in both disorders. Another key criterion for a distinct disorder is that it breeds true. If the syndrome diagnosis changed among members of the same family (where the mutation is obviously constant), it should raise the possibility that the allele at the primary locus does not specify a distinct syndrome and that there are major modifiers that determine a range of varying manifestations that make the syndrome difficult to describe as a distinct entity. In the case of PHS and GCPS, both disorders bred true, that is, no person with a diagnosis of PHS had been reported from a family with GCPS, and vice versa. At the time that the GLI3 mutations in PHS were discovered, all of the clinical data suggested that PHS and GCPS were distinct clinical entities. Importantly, the distinction of PHS and GCPS was qualitative, not quantitative. That is, no plausible spectrum could be hypothesized that put GCPS and PHS on a single scale of severity. Specifically, both GCPS and PHS had a wide spectrum of severity. Pallister-Hall syndrome varied from severe, neonatal lethal cases to mildly affected patients with insertional Polydactyly, asymptomatic hypothalamic hamartomas, and asymptomatic bifid epiglottis. Greig cephalopolysyndactyly syndrome was recognized early on to comprise a wide spectrum of severity, including patients with very mild manifestations.'^^ On the other end of the GCPS spectrum were patients with severe limb anomalies, craniofacial anomalies, CNS anomalies (agenesis of the corpus callosum), and mental retardation with or without seizures. Therefore, both disorders varied from mild to severe, but that variation did not encompass changes in the malformations that distinguish GCPS and PHS. On that basis, it was necessary to propose a biologic model that accounted for the generation of two distinct phenotypes.

Animal Models of GLI3 Phenotypes Because the human clinical data were limited,* insight into the mechanism of GLI3 activity was needed. One of the earliest insights was that the mouse mutant extra toes {XtJ) was very similar to human GCPS,^^ which was later confirmed by the determination that Gli3 was deleted in the XtJ mouse mutant."^^ A mouse model of PHS was also developed and it also demonstrated a qualitatively distinct set of malformations than those seen in the Xt mouse. A key insight was also gained from work on Drosophila melanogaster when it was demonstrated that CI, theflyhomologue of the human 6^7 genes, was subject to post-translational modification that converted ci from a transcriptional activator to a repressor.^ '^^ The pathway controlling ci processing is regulated by hedgehog (HH), theflyhomologue of the human hedgehog family that includes sonic hedgehog. Based on these data, a model was proposed whereby GLI3 was also post-translationally processed and had both activator and repressor functions. Furthermore, the mutations in patients with GCPS and PHS had to affect these two functions in distinct ways, in order to explain why both disorders had qualitatively distinct phenotypes. * Both GCPS and PHS are rare, probably affecting fewer than 1/100,000 persons.

Hedgehog-Gli Signaling in! Human Disease

132

Table 2. Distinct classes of mutations inPHSandCCPS

Translocations Large deletions Missense Splicing 5'frameshift 3'frameshift

GCPS

PHS

4 19 5 6 13 12

0 0 0 0 0 13

Currently published data show that the two disorders are associated with different classes of mutations (Table 2)7'^^'"^^'^^ Note that only one class of mutations is associated with PHS, truncating mutations 3' of the zinc finger binding domains. In contrast, mutations in patients with GCPS include translocations, large deletions,* missense, and frameshifi: mutations predominantly 5' of the zinc finger domains, but also in other areas of the gene.* Therefore, the range of mutations in GCPS is broader than in PHS and this observation is consistent with the hypothesis that GLI3 haploinsufficiency causes GCPS. However, the hypothesis that GLI3 haploinsufFiciency causes PHS is untenable, as no PHS patients have been described with mutations that cause haploinsufficiency. We proposed that mutations in GLI3 that cause PHS do so by generating a constitutive repressor that is independent of regulation by SHH. Numerous studies have shown that GLI3 (and GLI2, but interestingly not GLIl) are subjected to posttranslational processing and GLI3 has both activator and repressor activities. While these data substantially validate the bifimctional hypothesis, many questions remain. First are the so-called nonsyndromic Polydactylies. As noted above, clinical observations in the 1980s suggested that some individuals with GCPS have sufficiendy mild craniofacial manifestations that many observers would find it difficidt to recognize without detailed objective comparisons to unaffected relatives. Indeed, these features may be missed for two reasons. First, mild hypertelorism is considered a desirable trait (note that many models and actors have mild hypertelorism) and second, benign familial macrocephaly is one of the most common familial traits. Therefore, if a child has mild GCPS caused by a de novo mutation in GLI3^ general physicians and surgeons (the practitioners who would normally care for such a child) commonly miss the diagnosis of GCPS and instead diagnose the child with nonsyndromic Polydactyly. In such a case, the clinical geneticist may not get involved with the family and make a diagnosis of GCPS until that child has children of their own. For PHS the situation is less clear. As noted above, PHS does have a wide range of severity, including patients with Polydactyly, asymptomatic hypothalamic hamartomas (detectable only by cranial MRI), and asymptomatic bifid epiglottis (detectable by direct or indirect laryngoscopy**). We have evaluated a number of patients with apparendy nonsyndromic central Polydactyly who were diagnosed with mild PHS only after cranial MRI and laryngoscopy was performed and showed asymptomatic hypothalamic hamartoma or bifid epiglottis (unpublished data). In addition, some patients are diagnosed with PHS because they have postaxial Polydactyly (without central Polydactyly), because they are related to a person with PHS in a * Here we define "large" as being larger than the gene, which is --300 kb. * Unpublished data include more than 40 additional mutations that are consistent with these preliminary data. ** Remarkably, this lesion is often missed by anesthesiologists when these patients are intubated at the time of repair of their Polydactyly.

Human Correlates ofGLI3 Function

133

pattern consistent with autosomal dominant inheritance. The obvious implication is that there are likely to be individuals who have postaxial Polydactyly and asymptomatic hypothalamic hamartomas who are not related to someone with central Polydactyly. Such patients have syndromic Polydactyly, but unless they are carefully evaluated, would carry a diagnosis of nonsyndromic Polydactyly This brings to mind the report of families with so-called PAP-A (post axial Polydactyly type A, the "typ^ ^ " designation indicating that the extra digit is comparable in size to the normal fifth digit)^^ who have mutations in GLI3 that are similar to those in patients with PHS. Importandy, these individuals were not examined by cranial MRI or endoscopy so it is not known if they have asymptomatic hamartomas nor bifid epiglotti. Therefore, two hypotheses could explain the so-called nonsyndromic PAP-A patients with GLI3 mutations. First, such patients may have undiagnosed mild PHS. Second, it is possible that they are correcdy described as nonsyndromic and that PHS comprises a spectrum from severe, sporadic, neonatal presentations, through more typical patients, to mildly affected patients with small and asymptomatic hamartomas and finally, patients with nonsyndromic PAP-A. If the latter explanation is correct, other mechanisms must be invoked to explain the mildness of symptoms in the latter patients. One possibility is modifier genes. It is possible that variation in the expression of the normal and abnormal GLI3 alleles would affect the expression of the phenotype. The latter would be expected to be in cis with the mutations and thus be linked in the families, whereas the former would be expected to vary with the variant in the wild type allele. The are a number of coding polymorphisms in GZ/3, whose functional consequences are unknown, and the regulatory and splicing polymorphisms have not yet been studied. Thus, this area is ripe for further study. Another area of study is other classes of mutations. For example, as it is now known that GLI3 processing occurs, it is reasonable to expect that there could be amino acid substitution mutations in the gene that would render GLI3 resistant to processing. Such mutations should cause a phenotype distinct from PHS or GCPS, although the nature of this phenotype is difficult to predict. Another interesting class of mutations are those that cause amino acid substitutions that interfere with binding of GLI3 to the cytoplasmic complex. Such mutations might be nulls and thus phenotypically resemble GCPS, but it is also possible that they will cause constitutive GLI3 activator function, the phenotypic consequences of which are again difficult to predict. In this light, the recent report of acrocallosal syndrome caused by a point mutation in GLI3 raises another set of questions.^^ The overlap of GCPS and acrocallosal syndrome was recognized by clinicians before the GLI3 mutations were identified.^^ Part of this overlap is undoubtedly due to misclassification, primarily of sporadic cases. Acrocallosal syndrome is inherited in an autosomal recessive pattern and shares essentially all of the manifestations of GCPS except that agenesis of the corpus callosum, mental retardation, and seizures are common, if not universal. Prior to molecular diagnostics, if a sporadic patient presented with hypertelorism and preaxial polysyndactyly they were diagnosed with GCPS if they did not have mental retardation and seizures and acrocallosal syndrome if they did. In addition, the GCPS contiguous gene syndrome overlaps substantially with acrocallosal syndrome but is caused by large deletions in 7p that remove sufficient additional genes to cause CNS dysmyelination, mental retardation, and seizures (all of which are relatively nonspecific symptoms). ' Nevertheless, the data are convincing that acrocallosal syndrome is a distinct clinical entity that has a phenocopy caused by mutations in GLI3. In addition to the GCPS contiguous syndrome patients, one patient with acrocallosal syndrome and a point mutation in GLI3 has been described. No functional data on this mutant are available, but it may prove to be a dominant negative mutation of some kind. A final class of interesting patients is that with unclassified types of the oral-facial-digital syndrome (OFD). The OFD category describes a family of disorders with cleft palate, oral hamartomas, and frenulae, limb and other anomalies and currendy comprises ten apparently distinct phenotypes. There is an informal moratorium on designating additional subtypes of OFD because the boundaries of the entities are not clear and it is felt that molecular delineation will lead to a more rational distinction of the

134


various subtypes. Among patients with unclassified OFD, two patients have been described withfirameshifi;mutations in GLI3 (unpubUshed data). In summary, the cUnical and molecular analyses of patients with mutations in GLI3 have provided a wealth of insight into the biology of GLI3. In addition, these findings have result in direct benefits of improved diagnosis, treatment, and recurrence risk assessments for families affected by these disorders. For the fixture, it is clear that there is much yet to be studied and learned about the manifestations of GLI3 dysfiinction in humans.

Note

Added'in-Proof

A manuscript describing the genotype-phenotype correlation was published while this chapter was in production. Johnston JJ, Olivos-Glander I, Killoran C et al. Molecular and clinical analyses of Greig cephalopolysyndactyly and Pallister Hall syndromes: Robust phenotype prediction from the type and position of GLI3 mutations. Am J Hum Genet 2005 76:609-22.

References 1. Greig DM. Oxycephaly Edinb Med J 1926; 33:189-218. 2. Temtamy SA, McKusick VA. The genetics of hand malformations. Birth Defects Orig Artie Ser 1978; I4(i-xviii): 1-619. 3. Greig DM. Hypertelorism. Edinb Med J 1924; 31:560-593. 4. Gorlin RJ, Cohen Jr MM, Hennekam RCM. Greig cephalopolysyndactyly syndrome. Syndromes of the Head and Neck. 4th ed. New York: Oxford University Press, 2001:995-996. 5. Biesecker LG. GLI3 and the Pallister-Hall and Greig cephalopolysyndactyly syndromes. In: Epstein CJ, Erickson RP, Wynshaw-Boris A, eds. Inborn errors of development: The molecular basis of clinical disorders of morphogenesis. Oxford: Oxford University Press, 2004:257-264. 6. Motegi T, Ohuchi M, Ohtaki C et al. A craniosynostosis in a boy with a del(7)(pl5.3p21.3): Assignment by deletion mapping of the critical segment for craniosynostosis to the mid-portion of 7p21. Hum Genet 1985; 71:160-162. 7. Tommerup N, Nielsen F. A famiUal reciprocal translocation t(3;7) (p21.1;pl3) associated with the Greig polysyndactyly-craniofacial anomalies syndrome. Am J Med Genet 1983; 16:313-321. 8. Bacchichetti C, Artifoni L, Zanardo V. Deletions of the short arm of chromosome 7 without craniosynostosis. Clin Genet 1982; 21:348-349. 9. Vortkamp A, Gessler M, Grzeschik K-H. GLI3 zinc finger gene interrupted by translocations in Greig syndrome families. Nature 1991; 352:539-540. 10. Hall JG, Pallister SK, Clarren SK et al. Congenital hypothalamic hamartoblastoma, hypopituitarism, imperforate anus, and postaxial polydactyly-A new syndrome? Part I: Clinical, causal, and pathogenetic considerations. Am J Med Genet 1980; 7:47-7A. 11. Sills IN, Rapaport R, Robinson P et al. Familial Pallister-Hall syndrome: Case report and hormonal evaluation. Am J Med Genet 1993; 47:321-325. 12. Topf KF, Kletter GB, Kelch RP et al. Autosomal dominant transmission of the Pallister-Hall syndrome. J Pediatr 1993; 123:943-946. 13. Biesecker LG, Topf K, Graham Jr JM. FamiHal Pallister-Hall syndrome. J Med Genet 1994; 31:740. 14. Penman Splitt M, Wright C, Perry R et al. Autosomal dominant transmission of the Pallister-Hall syndrome. Clin Dysmorphol 1994; 3:301-308. 15. Sama A, Mason JD, Gibbin KP et al. Familial Pallister-Hall syndrome. J Med Genet 1994; 31:740. 16. Thomas HM, Todd PJ, Heaf D et al. Recurrence of PalUster-Hall syndrome in two sibs. J Med Genet 1994; 31:145-147. 17. Biesecker LG, Abbott M, Allen J et al. Report from the workshop on Pallister-Hall syndrome and related disorders. Am J Med Genet 1996; 65:76-81. 18. Kang S, Allen J, Graham Jr JM et al. Linkage mapping and phenotypic analysis of autosomal dominant Pallister-Hall syndrome. J Med Genet 1997; 34:441-446. 19. Kang S, Graham Jr JM, Olney AH et al. GLI3 frameshift mutations cause autosomal dominant PalHster-Hall syndrome. Nature Genet 1997; 15:266-268. 20. Baraitser M, Winter RM, Brett EM. Greig cephalopolysyndactyly: Report of 13 affected individuals in three families. CHn Genet 1983; 24:257-265. 21. Brueton L, Huson SM, Winter RM et al. Chromosomal localisation of a developmental gene in man: direct DNA analysis demonstrates that Greig cephalopolysyndactyly maps to 7pl3. Am J Med Genet 1988; 31:799-804.

Human

Correlates ofGLI3

Function

135

22. Vortkamp A, Franz T , Gessler M et al. Deletion of GLI3 supports the homology of the human Greig cephalopolysyndactyly syndrome (GCPS) and the mouse mutant extra toes (Xt). M a m m Genome 1992; 3:461-463. 23. Bose J, Grotewold L, Ruther U . Pallister-Hall syndrome phenotype in mice mutant for Gli3. H u m Mol Genet 2002; 11:1129-1135. 24. Aza-Blanc P, Kornberg T. Ci: A complex transducer of the Hedgehog signal. Trends Genet 1999; 15:458-462. 25. Aza-Blanc P, Ramfrez-Weber F-A, Laget M-P et al. Proteolysis that is inhibited by hedgehog targets C u b i t u s interruptus protein to the nucleus and converts it to a repressor. Cell 1997; 89:1043-1053. 26. Biesecker LG. Strike three for GLI3. Nature Genet 1997; 17:259-260. 27. Kalff-Suske M , Wild A, T o p p J et al. Point mutations throughout the GLI3 gene cause Greig cephalopolysyndactyly syndrome. H u m Mol Genet 1999; 8:1769-1777. 28. Sobetzko D , Eich G, Kalff-Suske M et al. Boy with syndactylies, macrocephaly, and severe skeletal dysplasiaL N o t a, new syndrome, but two^dominan^jnutations (GLI3 E543X and C O L 2 A 1 G973R) in the same individual. Am J Med Genet 2000; 90:239-T42; 29. Kalff-Suske M . Gene symbol: GLI3. Disease: Greig cephalopolysyndactyly syndrome. H u m Genet 2000; 107:203. 30. Kalff-Suske M . Gene symbol: GLI3. Disease: Pallister-Hall syndrome. H u m Genet 2000; 107:204. 3 1 . Kalff-Suske M , Paparidis Z, Bornholdt D et al. Gene symbol: GLI3. Disease: Pallister-Hall syndrome. H u m Genet 2004; 114:403. 32. Killoran CE, Abbott M, McKusick VA et al. Overlap of PIV syndrome, VACTERL and Pallister-Hall syndrome: Clinical and molecular analysis. Clin Genet 2000; 58:28-30. 33. Kroisel P M , Petek E, Wagner K. Phenotype of five patients with Greig syndrome and microdeletion of 7 p l 3 . A m J Med Genet 2 0 0 1 ; 102:243-249. 34. Wild A, Kalff-Suske M , Vortkamp A et al. Point mutations in h u m a n GLI3 cause Greig syndrome. H u m Mol Genet 1997; 6:1979-1984. 35. Williams P C , Hersh J H , Yen FF et al. Greig cephalopolysyndactyly syndrome: Altered phenotype of a microdeletion syndrome due to the presence of a cytogenetic abnormality. Clin Genet 1997; 52:436-441. 36. T u r n e r C , Killoran C , T h o m a s N S et al. H u m a n g e n e t i c disease caused b y de n o v o mitochondrial-nuclear D N A transfer. H u m Genet 2003; 112:303-309. 37. Radhakrishna U, Bornholdt D , Scott H S et al. T h e phenotypic spectrum of GLI3 morphopathies includes autosomal dominant preaxial Polydactyly type-IV and postaxial Polydactyly type- A/B; N o phenotype prediction from the position of GLI3 mutations. Am J H u m Genet 1999; 65:645-655. 38. Wagner K, Kroisel P M , Rosenkranz W . Molecular and cytogenetic analysis in two patients with microdeletions of 7p and Greig syndrome: Hemizygosity for P G A M 2 and T C R G genes. Genomics 1990; 8:487-491. 39. Pelz L, Kruger G, Gotz J. T h e Greig cephalopolysyndactyly syndrome. Helv Paediatr Acta 1986; 41:381-382. 40. Pettigrew AL, Greenberg F, Caskey C T et al. Greig syndrome associated with an interstitial deletion of 7p: Confirmation of the localization of Greig syndrome to 7 p l 3 . H u m Genet 1991; 87:452-456. 4 1 . Driess S, Freese K, Bornholdt D et al. Gene symbol: GLI3. Disease: Greig cephalopolysyndactyly syndrome. H u m Genet 2003; 112:103. 42. Freese K, Driess S, Bornholdt D et al. Gene symbol: GLI3. Disease: Pallister-Hall syndrome. H u m Genet 2 0 0 3 ; 112:103. 43. Friez MJ, Stevenson RE. Pallister-Hall syndrome. Proc Greenwood Genet Cen 2000; 19:51-55. 44. Galasso C, Scire G, Fabbri F et al. Long-term treatment with growth hormone improves final height in a patient with Pallister-Hall syndrome. A m J Med Genet 2 0 0 1 ; 99:128-131. 45. Johnston JJ, Olivos-Glander I, Turner J et al. Clinical and molecular delineation of the Greig cephalopolysyndactyly contiguous gene deletion syndrome and its distinction from acrocallosal syndrome. A m J Med Genet 2003; 123A:236-242. 46. N g D , Johnston JJ, Turner J T et al. Gonadal mosaicism in severe Pallister-Hall syndrome. Am J Med Genet 2004; 124A:296-302. 47. Bianchi D W , Cirillo-Silengo M , Luzzatti L et al. Interstitial deletion of the short arm of chromosome 7 without craniosynostosis. Clin Genet 1981; 19:456-461. 48. Marks K, Hill L, Chitham RG et al. Interstitial deletion of chromosome 7p detected antenatally. J Med Genet 1985; 22:316-318.

136

Hedgehog-Gli

Signaling in Human

Disease

49. Debeer P, Peeters H , Driess S et al. Variable phenotype in Greig cephalopolysyndactyly syndrome: Clinical and radiological findings in 4 independent families and 3 sporadic cases with identified GLI3 mutations. A m J M e d Genet 2003; 120A:49-58. 50. Kremer S, Minotti L, Thiriaux A et al. Epilepsy and hypothalamic hamartoma: Lx)ok at the hand Pallister-Hall syndrome. Epileptic Disord 2003; 5:27-30. 5 1 . Bai C B , Stephen D , Joyner AL. All mouse ventral spinal cord patterning by hedgehog is Gli dependent and involves an activator function of Gli3. Dev Cell 2004; 6:103-115. 52. Li Y, Zhang H , Choi SC et al. Sonic hedgehog signaling regulates Gli3 processing, mesenchymal proliferation, and differentiation during mouse lung organogenesis. Dev Biol 2004; 270:214-231. 53. Persson M , Stamataki D , te Welscher P et al. Dorsal-ventral patterning of the spinal cord requires Gli3 transcriptional repressor activity. Genes Dev 2002; 16:2865-2878. 54. Shin S, Kogerman P, Lindstrom E et al. GLI3 mutations in human disorders mimic Drosophila Cubitus interruptus protein functions and localization. Proc Natl Acad Sci USA 1999; 96:2880-2884. 55. W a n g B, Fallon J, Beachy P. Hedgehog-regulated processing of Gli3 produces an anterior/poterior repressor gradient in the developing vertebrate limb. Cell 2000; 100:423-434. 56. Ruiz i Altaba A. Gli proteins encode context-dependent positive and negative functions: Implications for development and disease. Development 1999; 126:3205-3216. 57. Asch AJ, Myers GJ. Benign familial macrocephaly: Report of a family and review of the literature. Pediatrics 1976; 57:535-539. 58. Radhakrishna U, Wild A, Grzeschik K-H et al. Mutation in GLI3 in post-axial Polydactyly type A. Nature Genet 1997; 17:269-270. 59. Elson E, Perveen R, Donnai D et al. De novo GLI3 mutation in acrocallosal syndrome: Broadening the phenotypic spectrum of GLI3 defects and overlap with murine models. J Med Genet 2002; 39:804-806. 60. Gorlin RJ, Cohen Jr M M , Hennekam R C M . T h e oral facial digital syndromes. In: GorUn RJ, Cohen Jr M M , Hennekam R C M , eds. Syndromes of the head and neck. 4th ed. Oxford: Oxford University Press, 2 0 0 1 .

CHAPTER 11

From Oligodactyly to Polydactyly. Role of Shh and Gli3 in Limb Morphogenesis Chin Chiang* Abstract

S

ecreted molecules encoded by the Hedgehog (Hh) gene family have emerged as key signals in regulating the growth and patterning of invertebrate and vertebrate embryos. One of the most prominent features among Hh members is thought to reside in their ability to impose distinct cell fates in a concentration-dependent manner. This ability is highlighted by the critical and indispensable role of Sonic hedgehog (Shh) signaling in specifying the anterior-posterior polarity of the embryonic limb. Alteration of Shh expression and signaling activity can lead to profound developmental abnormalities in digit numbers and identity in mice and humans. In this chapter, we discuss the Shh regulatory mechanism that establishes the anterior-posterior polarity of the limb and how misregulation of this mechanism can lead to severe limb malformations in humans.

Introduction All tetrapod limbs consist of three skeletal segments (Fig. 1), with the proximal stylopod (humerus in forelimb or femur in hindlimb) articulating a pair of zeugopods (radius/ulna or tibia/fibula), followed by the autopod (wrist/hand or ankle/foot). These complex structures are products of the outgrowth and patterning of paired forelimb and hindlimb buds which originate from the lateral plate mesoderm by inductive signals. ' The anterior-posterior (A/P) asymmetry of the normal five-digit pentadactyl limb, as reflected by the thumb and small finger, is controlled by a group of specialized mesodermal cells located at the posterior margin of the limb bud referred to as the zone of polarizing activity (ZPA) (Fig. 1). Grafi:ing of the entire ZPA to the anterior mesoderm results in a complete mirror-image duplication of digits while partial ZPA grafts result in duplication of a subset of digits. The ability of cells in the ZPA to progressively specify an increase in the number of digits with more posterior identities suggested the existence of a diffusible molecule (morphogen) in the establishment of A/P polarity. A major advance in understanding the molecular control of A/P limb patterning came from the discovery of Shh, a secreted signaling molecule expressed in the ZPA. Shh is capable of inducing supernumerary digits and modifying digit identity when misexpressed anteriorly^'^ (Fig. 1). Analysis of mouse embryos with a targeted deletion of Shh indicated an absolute requirement of Shh for ZPA function.^ Additionally, the Shh mutant exhibited truncations of distal skeletal elements in the forelimbs while its hindlimbs displayed a single un-ossified but identifiable digit 1^'^^ (Fig. 3). These studies provide an important foundation for elucidating the molecular mechanism of human congenital limb malformations which encompass a broad *Chin Chiang—Department of Cell and Developmental Biology, Vanderbilt University Medical Center, 465 21^* Ave. South, Nashville, Tennessee 37232, U.S.A. Email: [email protected]


138


Figure 1. Shh mediates the zone of polarizing activity (ZPA) in the limb. A) the limb consists of three segments: the stylopodium (st), zeugopodium (ze) and autopodium (au), in a proximal to distal order as shown for the mouse forelimb. Shh expression in the ZPA at the posterior (P) margin of the limb bud is necessary for anterior-posterior (A/P) patterning of distal skeletal elements. B) Grafting of the ZPA or Shh-soaked beads to the anterior margin of the limb mesoderm (red) induces supernumerary digits (4-3-2-2-3-4) in chick.^ Note that arrow in b highlights the duplicated radius (r). Abbreviations: h, humerus; u, ulna and distal is to the right. A color version of this figure is available online at http://www.Eurekah.com. range of limb phenotypes from oligodactyly to Polydactyly, with fewer or more than the normal number of digits, respectively.

Shh Expression and the Specification of Digit Identity Restricted expression of Shh in the ZPA permits the generation of a Shh concentration gradient along the A/P axis of the developing limb bud. Analysis of several preaxial polydactylous (PPD) mutants provided evidence for the existence of a negative regulatory program that restricts Shh expression to the posterior margin of the limb bud. A common feature shared by these PPD mouse mutants is ectopic Shh expression in the anterior margin of the limb mesoderm, consistent with the generation of additional digits in the anterior side of the limb. Several of these mutants are now molecidarly characterized including Strongs Itixoid (lst)y Sasquatch (Ssq), Hemimelic extratoes (Hx), andM 10081}^'^^ The 1st gene, which encodes a paired-type homeodomain protein, was previously described as the Aristaless-like4 (Alx4) gene.^ In AIx4 mutants, the critical arginine residue within the first helix of the homeodomain is substituted by glutamic acid. This mutation is predicted to result in AlxA loss-of-function due to disruption of DNA binding at target sites. In addition to developmental defects in the limb, Abc4 mutants also display craniofacial and hair defects and failure of testicular descent into the scrotum. ^^ Alx4 is expressed in the anterior mesoderm opposite the ZPA, consistent with its role in repressing Shh expression in the anterior

From Oligodactyly to Polydactyly

139

Ssq insertion

LMBRI El

SHH 23

16 17

i—I /

.\

Acheiropodia/ /

.k GCPS

PHS PAP-A

Q496X

>k

Nonsense

671 (-1) , 674 (-1) Frameshift 764 (-1)

Frameshift

GCPS

R792X

Nonsense

PHS

E1147X

Nonsense

GCPS I

AT

1168(-1)

Frameshift

Figure 5. Examples of GLI3 mutations in GCPS, PHS and PAP. Schematic diagram of GLI3 and GLI3R proteins with several known functional domains such as the suppressor of fused [Su(fu)] binding domain, zincfingerdomain (ZFD) and GBP binding domain. The types of mutations are indicated as nonsense, missense orfi-ameshift.The number designation for each GLI3 mutation specifies the altered amino acid residue. Data are fromreferences42, 43, 46 and Al. expression profiles that are associated with particular digit identities. Thus, a higher Gli3A:Gli3R threshold level will generate digits with posterior identities while a lower threshold level will generate digits of anterior identities. This model is consistent with the finding that homozygous GliJ' or Shh' ';Gli3'' limbs exhibit a complete loss of wild-type digit identities. Further support for this model comes from analysis of talpid2 {tO^ chick mutant limbs which show uniform expression of Shh target genes and high Gli3A:Gli3R levels across the A/P axis. Accordingly, ta2 limbs exhibit Polydactyly and uniform posterior digit identities. ^

Congenital Limb Malformations Are Associated with Gli3 Mutations Mutations in human GLI3 have been identified in Greig Cephalopolysyndactyly syndrome (GCPS),^2 Pallister-Hall syndrome (PHS)^^ and postaxial Polydactyly type A (PAP-A).^^ GCPS and PHS are autosomal dominant conditions displaying defects in growth and development of the head, limb and visceral organs, while PAP-A affects only the development of the limb. The hallmarks of GCPS are preaxial Polydactyly with facial anomalies such as hypertelorism and frontal bossing. These characteristics are distinct from PHS which displays central or postaxial type Polydactyly, hypothalamic hamartoblastoma and, occasionally, neural midline defects such as abnormal pituitary and callosal agenesis. The clinical manifestations of these three syndromes are quite distinct, but all exhibit Polydactyly and syndactyly. It has been proposed that defined truncations of the GLI3 protein are associated with specific phenotypes ^ (Fig. 5). The dominant preaxial Polydactyly phenotype in GCPS patients and in mouse Xtj^ is caused by genetic mutations that disrupt the function of one Gli3 allele. In contrast, PHS and PAP-A syndromes are associated with mutations predicted to generate altered Gli3 proteins with truncations carboxy-terminal to the zinc finger domain (ZFD) (see Fig. 1). These truncated proteins will not be regulated by Shh at the level of processing. Thus,


143

the distinct characteristics of PHS and PAP-A are possibly due to enhanced production of GLI3R proteins, whereas GCPS phenotypes are caused by GLI3 hapioinsufFiciency (Fig. 1 A, structurephenotype model). Since individuals with PAP-A have a milder phenotype than those with PHS, this observation suggests that the sites mutated in PAP-A may attenuate repression by truncated GLI3 proteins. However, recent analysis of additional families displaying these syndromes suggests that the molecular mechanism leading to a clinical manifestation may be more complex than predicted based on the structurephenotype model. For example, familial patients carrying a nonsense mutation in codon 792 (R792), predicted to harbor a GLI3 truncation right after the zinc-finger domain, were diagnosed with GCPS instead of PHS (Fig. IB). These observations suggest that the expression of distinct phenotypic traits in these syndromes likely reflects how a particular mutation affects GLI3 transcriptional properties.

TZoncIusion Polydactyly is one of the most common types of limb malformations in humans. Analyses of mouse mutants have provided insights into the molecular mechanism leading to Polydactyly as observed in humans. The emergence of two conceptual classes of Polydactyly provides a link between developmental mechanisms and phenotypic traits.^^ Group 1 mutations result from ectopic Shh expression in the anterior limb bud mesoderm but do not disrupt Gli3 function. Group 1 mutation is characterized by preaxial Polydactyly with mirror image digit duplications, which is exemplified by mouse mutants such as Ist^ Ssq, Hx, Ml00081 and the human PPD locus. Additional mouse mutants such as Recombination induced mutant 4 (Rim4), X-linked Polydactyly (Kpt) and luxate (Ix)^^' also belong to Group 1, and elucidating the molecular nature of these mutants will likely provide mechanistic insights into the negative regulatory program that controls Shh expression in the limb. In contrast. Group 2 mutations disrupt Gli3 fiinction by a Shh-independent mechanism thereby generating a spectrum of supernumerary digits. Group 2 polydactylous limbs lack mirror image duplications and their supernumerary digits are always syndactylous. The phenotypic variations in GCPS, PHS and PAP-A syndromes likely reflect how a particular Gli3 mutation affects its transcriptional activities and target gene expression. Functional characterization of these mutant proteins in mice will likely provide novel insights into the developmental mechanisms of Polydactyly in humans.

References 1. Tickle C. Patterning systems—from one end of the limb to the other. Dev Cell 2003; 4(4):449-458. 2. Martin G R . T h e roles of FGFs in the early development of vertebrate limbs. Genes Dev 1998; 12:1571-1586. 3. Saunders J W , Gasseling M . Ectodermal-mesenchymal interaction in the origin of limb symmetry. In: Fleischmayer R, Billingham RE, eds. Epithelial-mesenchymal interaction. Baltimore: Williams and Wilkins, 1968:78097. 4. Tickle C. T h e number of polarizing region cells required to specify additional digits in the developing chick wing. Nature 1981; 289:295-298. 5. Chang D T , Lopez A, von Kessler D P et al. Products, genetic linkage and limb patterning activity of a murine hedgehog gene. Development 1994; 120:3339-3353. 6. Lopez-Martinez A, Chang D T , Chiang C et al. Limb-patterning activity and restricted posterior localization of the amino-terminal product of Sonic hedgehog cleavage. Current Biology 1995; 5:791-796. 7. Riddle R D , Johnson RL, Laufer E et al. Sonic hedgehog mediates the polarizing activity of the ZPA. Cell 1993; 75:1401-1416. 8. Chiang C , Litingtung Y, Harris M P et al. Manifestation of the limb prepattern: Limb development in the absence of sonic hedgehog function. Dev Biol 2 0 0 1 ; 236(2):421-435. 9. Kraus P, Fraidenraich D , Loomis CA. Some distal limb structures develop in mice lacking Sonic hedgehog signaling Mech Dev 2 0 0 1 ; 100(l):45-58. 10. Lewis P M , D u n n M P , M c M a h o n JA et al. Cholesterol modification of sonic hedgehog is required for long-range signaling activity and effective m o d u l a t i o n of signaling by P t c l . Cell 2 0 0 1 ; 105(5):599-612.

144

Hedgehog-Gli

Signaling in Human

Disease

11. Q u S, Tucker SC, Ehrlich JS et al. Mutations in mouse Aristaless-like4 cause Strong's luxoid Polydactyly. Development 1998; 125(14):2711-2721. 12. Lettice LA, Heaney SJ, Purdie LA et al. A long-range Shh enhancer regulates expression in the developing limb and fin and is associated with preaxial Polydactyly. H u m Mol Genet 2003; 12(14):1725-1735. 13. Sagai T, Masuya H , T a m u r a M et al. Phylogenetic conservation of a limb-specific, cis-acting regulator of Sonic hedgehog ( Shh). M a m m Genome 2004; 15(l):23-34. 14. Q u S, Tucker SC, EhrUch JS et al. Mutaions in mouse aristaless-like4 cause strong's luxoid Polydactyly. Development 1998; 125:2711-2721. 15. Q u S, Niswender K D , Ji Q et al. Polydactyly and ectopic ZPA formation in Alx-4 mutant mice. Development 1997; 124(20):3999-4008. 16. te Welscher P, Zuniga A, Kuijper S et al. Progression of vertebrate limb development through SHH-mediated counteraction of GLI3. Science 2002; 298(5594):827-830. 17. Lettice LA, Horikoshi T , Heaney SJ et al. Disruption of a long-range cis-acting regulator for Shh causes preaxial Polydactyly. Proc Natl Acad Sci USA 2002; 9 9 ( l l ) : 7 5 4 8 - 7 5 5 3 . 18. Clark R M , Marker P C , Kingsley D M . A novel candidate gene for mouse and human preaxial Polydactyly with altered expression in limbs of Hemimelic extra-toes mutant mice. Genomics 2000; 67(1): 19-27. 19. Hill RE, Heaney SJ, Lettice LA. Sonic hedgehog: Restricted expression and limb dysmorphologies. J Anat 2003; 202(l):13-20. 20. Maas SA, Fallon JF. Isolation of the chicken Lmbrl coding sequence and characterization of its role during chick Hmb development. Dev Dyn 2004; 229(3): 520-528. 2 1 . Castilla EE, Lugarinho da Fonseca R, da Graca Dutra M et al. Epidemiological analysis of rare Polydactylies. Am J Med Genet 1996; 65(4):295-303. 22. Temtamy SA, McKusick VA. T h e genetics of hand malformations. Birth Defects Orig Artie Ser 1978; I4(3):(i-xviii), 1-619. 2 3 . Heutink P, Zguricas J, van Oosterhout L et al. T h e gene for triphalangeal t h u m b maps to the subtelomeric region of chromosome 7q. Nat Genet 1994; 6(3):287-292. 24. Tsukurov O , Boehmer A, Flynn J et al. A complex bilateral polysyndactyly disease locus maps to chromosome 7q36. N a t Genet 1994; 6(3):282-286. 25. Zguricas J, Heus H , Morales-Peralta E et al. Clinical and genetic studies on 12 preaxial Polydactyly families and refinement of the localisation of the gene responsible to a 1.9 cM region on chromosome 7q36. J Med Genet 1999; 36(l):32-40. 26. lanakiev P, van Baren MJ, Daly MJ et al. Acheiropodia is caused by a genomic deletion in C7orf2, the human orthologue of the Lmbrl gene. Am J H u m Genet 2 0 0 1 ; 68(l):38-45. 27. Clark RM, Marker PC, Roessler E et al. Reciprocal mouse and human limb phcnotypes caused by gain- and loss-of-function mutations affecting L m b r l . Genetics 2 0 0 1 ; 159(2):715-726. 28. Ingham PW, M c M a h o n AP. Hedgehog signaling in animal development: Paradigms and principles. Genes Dev 2 0 0 1 ; 15(23):3059-3087. 29. M o R, Freer AM, Zinyk D L et al. Specific and redundant functions of Gli2 and Gli3 zinc finger genes in skeletal patterning and development. Development 1997; 124(1): 113-123. 30. Park HL, Bai C, Piatt KA et al. Mouse Glil mutants are viable but have defects in S H H signaling in combination with a Gli2 mutation. Development 2000; 127(8): 1593-1605. 3 1 . Johnson DR. Extra-toes: A new mutant gene causing multiple abnormalities in the mouse. J Embyol Exp Morph 1967; 17:543-581. 32. H u i C C , Joyner AL. A mouse model of greig cephalopolysyndactyly syndrome: The extra-toesj mutation contains an intragenic deletion of the Gli3 gene. Nature Genetics 1993; 3(3):24l-246. 33. Masuya H , Sagai T, Moriwaki K et al. Multigenic control of the localization of the zone of polarizing activity in limb morphogenesis in the mouse. Dev Biol 1997; 182(1):42-51. 34. Buscher D , Ruther U. Expression profile of Gli family members and Shh in normal and mutant mouse limb development. Developmental Dynamics 1998; 211(l):88-96. 35. Litingtung Y, Dahn R D , Li Y et al. Shh and GH3 are dispensable for Hmb skeleton formation but regulate digit number and identity. Nature 2002; 4l8(6901):979-983. 36. Dai P, Akimaru H , Tanaka Y et al. Sonic Hedgehog-induced activation of the Glil promoter is mediated by GLI3. J Biol Chem 1999; 274(12):8143-8152. 37. Ruiz i Altaba A. Gli proteins encode context-dependent positive and negative functions: Implications for development and disease. Development 1999; 126(14):3205-3216. 38. Sasaki H , Nishizaki Y, H u i C et al. Regulation of Gli2 and Gli3 activities by an amino-terminal repression domain: Implication of Gli2 and Gli3 as primary mediators of Shh signaling. Development 1999; 126(17):3915-3924.


145

39. Shin SH, Kogerman P, Lindstrom E et al. GLI3 mutations in human disorders mimic Drosophila cubitus interruptus protein functions and localization. Proc Natl Acad Sci USA 1999; 96(6):2880-2884. 40. Wang B, Fallon JF, Beachy PA. Hedgehog-regulated processing of Gli3 produces an anterior/posterior repressor gradient in the developing vertebrate limb. Cell 2000; 100(4):423-434. 41. Caruccio NC, Martinez-Lopez A, Harris M et al. Constitutive activation of sonic hedgehog signaling in the chicken mutant talpid(2): Shh-independent outgrowth and polarizing activity [In Process Citation]. Dev Biol 1999; 212(1):137-149. 42. Wild A, KalflF-Suske M, Vortkamp A et al. Point mutations in human GLI3 cause Greig syndrome. Hum Mol Genet 1997; 6(11): 1979-1984. 43. Kang S, Graham Jr JM, Olney AH et al. GLI3 frameshift mutations cause autosomal dominant PalHster-Hall syndrome. Nat Genet 1997; 15(3):266-268. 44. Radhakrishna U, Wild A, Grzeschik KH et al. Mutation in GLI3 in postaxial Polydactyly type A [letter]. Nat Genet 1997; 17(3):269-271. 45. Biesecker LG. Strike three for GLI3. Nature Genetics 1997; 17(3):259-260. 46^âdhalcrishnaXJ^ôrnhnldt D^^ScotiLJJS^^r al. The Phenotypic spectrum of GLI3 morphopathies includes autosomal dominant preaxial Polydactyly type-IV and postaxial Polydactyly type- A/B; no phenotype prediction from the position of GLI3 mutations. Am J Hum Genet 1999; 65(3):645-655. A7. KalfF-Suske M, Wild A, Topp J et al. Point mutations throughout the GLI3 gene cause Greig cephalopolysyndactyly syndrome. Hum Mol Genet 1999; 8(9): 1769-1777. 48. Debeer P, Peeters H, Driess S et al. Variable phenotype in Greig cephalopolysyndactyly syndrome: Clinical and radiological findings in 4 independent families and 3 sporadic cases with identified GLI3 mutations. Am J Med Genet 2003; 120A(l):49-58. 49. Masuya H, Sagai T, Wakana S et al. A duplicated zone of polarizing activity in polydactylous mouse mutants. Genes Dev 1995; 9:1645-1653.

CHAPTER 12

The Genetics of Indian Hedgehog M. Elizabeth McCready and Dennis E. Bulman*

T

he hedgehog proteins are highly conserved between species and act as key morphogens during development, along with other molecules such as fibroblast growth factors (FGFs), bone morphogenetic proteins (BMPs) and Hox proteins. In mammals there are three known family members, sonic hedgehog (Shh), desert hedgehog (Dhh) and indian hedgehog (Ihh). While these proteins are highly homologous and can substitute for one another in a variety of functional assays,^ they elicit distinct biological responses in vivo and in vitro.'^ Studies have shown that Ihh is critical for proper development of several tissues including the pancreas,^ and the endochondral skeleton.^' '^ There is also evidence that it is involved in hematopoiesis, vasculogenesis and colonocyte differentiation/ Thus far, mutations of the /////gene have been identified in two hereditary conditions, brachydactyly type Al (BDAl)^'^^ and acrocapitofemoral dysplasia (ACFD)/^ Surprisingly, despite the range of tissues that IHH modulates developmental and differentiation processes of, the only system that appears to be affected in these conditions, is the skeleton.^' Normal bone development can occur by one of two processes, intramembranous, or endochondral ossification. The first process involves direct ossification of mesenchymal tissue at the site of the future bone and occurs primarily in the flat bones of the skull. Endochondral ossification, by contrast, requires a cartilage intermediate and is the primary process by which bone development occurs in the axial and appendicular skeletons. As the limb grows along the proximodistal axis, mesenchymal cells begin to condense to form the basic shapes of the future bone. At the core of these condensations, cells differentiate into chondrocytes and start to secrete a cartilaginous matrix (reviewed in re£ 13). Periphery cells also differentiate and secrete a perichondrial sheath that surrounds the cartilage anlagen. Since bone does not have the capacity to grow, formation of a cartilaginous structure is essential for skeletal growth prior to bone deposition. Defects in proper cartilage analgen formation, or in mesenchymal condensation, can result in shortened, abnormally shaped or absent bones. Shordy after formation of the cartilage analgen, cells in the center of the structure exit the cell cycle, and mature into hypertrophic chondrocytes. The terminally differentiated chondrocytes finally die through apoptosis and the calcified matrix is replaced with trabecular bone. As the process of hypertrophy extends outwards from the center of the bone, the zone of proliferating chondrocytes is restricted to the apical ends. Secondary ossification centers may form at the distal ends of the bone, leaving a small region between the primary and secondary ossification center that continues to proliferate. This region is called the epiphyseal growth plate and is the only portion of the skeletal element that retains growth potential. The cells within this region arrange themselves in columns including a zone of resting chondrocytes followed by zones of proliferating, prehypertrophic, hypertrophic and terminally differentiated chondrocytes. The conversion of calcified cartilage to trabecular bone matrix continues until adolescence when the epiphyses close and the growth plate is finally replaced by bone. *Corresponcling Author: Dennis E. Bulman—Ottawa Health Research Institute, 501 Smyth Road, Ottawa, Ontario, Canada K1 H 8L6. Email: [email protected]


The Genetics of Indian Hedgehog

147

Within the growth plate, the transition of chondrocytes from the proUferative to terminally differentiated states is a highly regulated process. Maintenance of appropriate numbers of cells within each fraction is necessary to ensure normal growth and development. Ihh is a key modulator of proliferation and differentiation within the growth plate. Mice homozygous for an Ihh null allele {Ihh''' mice) show a reduced capacity for chondrocyte proliferation and a shift in the fraction of hypertrophic chondrocytes towards the axial ends. The absence of a bone collar in these mice further demonstrates a role for Ihh during osteogenesis, although it does not appear necessary for initiation of osteoblast differentiation. ' Finally, it has been shown that Ihh can directly promote chondrocyte proliferation and may be necessary for driving growth along the longitudal axis of the bone.^^

Brachydactyly Type A l The brachydactylies are a group of congenital malformations that affect normal bone development resulting m shortened digits. Thq^ are categorized into a number of different types based on the bones and digits involved. ' Brachydactyly type Al (BDAl) is characterized by bilateral shortness of the middle phalanges in digits two through five that is at least 2 standard deviations below normal. The observed phenotype is variable and can range in severity from a mild variant, in which the shortness of the middle phalanges is more apparent than in other digital bones, to a severe variant, in which the middle phalanges are absent or fused to the terminal phalanges. In both mild and severe forms of BDAl, the shortening is most prominent in digits two and five, followed by digits four, then three. Similarly, if terminal symphalangism is present, it is more common in digits two and five than four and three. Although shortening of the middle phalanges is the hallmark feature of BDAl a number of other skeletal features have been reported including short stature, ' radial/ulnar clinodactyly, ' shortening of the first, third,^^'^^ or fifth metacarpal, and premature epiphyseal closure.^^'^^'^ '^^ In a comprehensive analysis of affected individuals from an American family with BDAl, Haws and McKusick further described sloping of the distal end of the radius, an absence of the styloid process of the ulna, abnormally shallow acetabulum and glenoid fossa, and an elliptical, rather than round, humerous head.^ Incidents of scoliosis and BDAl have also been reported in at least three famiUes.^^'^^'^^ BDAl has the distinction of being the first human trait described in terms of autosomal dominant Mendelian inheritance.^^ Despite this, little was known about the genes involved until recently. In 2000, Yang et al described linkage of BDAl to chromosome 2q35-36 in two Chinese families with the trait.^^ This led to the subsequent identification of/////mutations at nucleotide positions 283 and 391 of the coding sequence in families I and II, respectively.^ An additional mutation at nucleotide position 300 was identified in a third family, further supporting a role for IHH in the pathophysiology of BDAl. The three mutations were predicted to cause amino acid substitutions at Glu95 (Glu^Lys), AsplOO (Asp->Glu) and Glul31 (Glu^Lys). Since the initial report of/////mutations, two additional allelic variants have been identified in individuals with BDAl .^'^ ^ The first, a G ^ A transition at nucleotide position 298, was identified in descendents of some of the earliest reports of BDAl inheritance, '^^'^^ and is predicted to cause an aspartic acid to asparagine amino acid substitution at codon 100. It has since been identified in four additional kindreds and likely represents an ancestral mutation (unpublished observation and ref 10). The second variant, an A ^ G transition was identified at nucleotide position 284 of the IHH coding sequence, in an American family of Scandinavian descent.^^ It is predicted to cause a glutamate to glycine amino acid substitution at codon 95. The AsplOOAsn and Glu95Gly variants both cause substitutions of identical amino acids to those described in the first report of BDAl mutations, although they result in different substitutions.^'^'^^ This suggests that these two amino acids may be hot spots for BDAl mutations and that they are likely important for normal IHH function during endochondral ossification.^ Although the tertiary structure of the human IHH protein has not yet been characterized, the crystal structure of Shh predicts that Glu95, AsplOO, and Glul31 are closely situated to

148


Figure 1. Tertiary structure of murine Shh N-terminal signalling domain indicating location of amino acids mutated in human disease. Based on high degrees of similarity between the murine Shh and human hedgehog N-terminal domains, the tertiary structure of murine Shh has been used to model the locations of disease-causing mutations reported in three human diseases, BDAl, ACFD, and holoprosencephaly. The latter disease is caused by mutations of the 5////gene. Green, red and yellow molecules indicate the position of amino acids that are mutated in BDAl, ACFD and holoprosencephaly, respectively. (Reproduced with permission from Hellemans J, Coucke PJ, Geidion A. Am J Hum Genet 2003; 72:1040-1046, University of Chicago Press. ©2003 by The American Society of Human Genetics.) one another in an external groove of the protein^ ^ in an area known to facilitate binding of the hedgehog proteins to the Patched ( P T C H ) receptor (see Fig. 1).^'32,33 Q^^ ^^ ^ predicted that mutations at these sites may alter I H H function by either disrupting interactions with it natural receptor(s), or by enhancing its ability to interact with alternate receptors. Attenuation of IHFi signaling may further decrease the growth potential of developing bones by reducing the number of proliferating chondrocytes, and/or increasing the number of cells differentiating from the proliferative to hypertrophic stages. As the middle phalanges are the last phalanges to ossify, they may be more susceptible to reduced chondrocyte numbers than other phalanges.^^ Mutations of the / / / / / g e n e account for only a subset of the B D A l incidents described to date, indicating that the trait is genetically heterogenous.^^'^ This is further supported by the identification of a second locus on chromosome 5pl3.3-13.2.^ A third locus may also exist, based on the identification of an American family with B D A l that lacked detectable IHH mutations in the coding region, and whose trait was not linked to the chromosome 5p locus. T h e identification of B D A l mutations at loci other than / / / / / i s important as this information may contribute significantly to our understanding of endochondral ossification in humans.

The Genetics of Indian Hedgehog

149

Characterization of alternate BDAl loci may also provide information about the IHH signaling pathway, or parallel biological pathways.

Acrocapitofetnoral Dysplasia (ACFD) ACFD is an autosomal recessive skeletal dysplasia characterized by short stature, brachydactyly (short middle and distal phalanges), a narrow thorax, and a relatively large head. Cone-shaped/ abnormal epiphyses and premature epiphyseal closure occur in the hands, the proximal portion of the femur and to a variable degree in the shoulders, knees and ankles of affected individuals.^^ The most consistent finding among individuals diagnosed with ACFD, is a round, egg-shaped femoral head with a shallow, almost absent femoral neck.^^ ACFD was first described in March 2003, in two consanguineous families with five affected individuals. * Homozygosity mapping of individuals from these families was used to define a 5 cM critical region on chromosome 2q35-q36. Using a positional candidate gene approach, the same group subsequently identified a C-^T transition at nucleotide 137 of the IHH gene (P46L) in family one, and a T - ^ C transition at nucleotide 569 of the same gene (V190A) in family two. Both mutations caused amino acid substitutions in the amino terminal portion of the protein responsible for local and long range signaling. ^' '^^ Hellemans et al hypothesized that the mutations cause ACFD by diminishing IHH signaling and thus increasing the rate of chondrocyte differentiation.^"^ Despite being heterozygous for the above mutations, the parents of each affected individual did not show signs of ACFD. The authors did note, however, that the parents from family one had short middle phalanges (up to -2.5 SD in digits 2-4 and up to -5 SD in digit 5), which were neither rudimentary nor fused to the terminal phalanges.^"^ The parents from family two were asymptomatic except for mild shortening of the metacarpals and proximal phalanges (> 2 SD) in two individuals.

Comparison of Human IHH Disorders As BDAl and ACFD are allelic, we may expect to observe some phenotypic similarities between the two conditions. The most consistent similarity between individuals affected with each condition, is brachydactyly involving the middle phalanges. Short stature and scoliosis have also been reported in association with both conditions, '^^''^^''^'^'^ '^^ but do not occur consistendy in all cases. Skeletal surveys are seldom reported in case studies of BDAl, making it difficult to compare other phenotypes between the two traits. However, in 1963, Haws and McKusick^^ published an extensive description of three individuals with BDAl, who were descendants from the first kindred used to demonstrate Mendelian dominant inheritance in humans. In addition to analyses of the hands and feet. Haws and McKusick provided detailed descriptions of malformations to the distal ends of the radius, the ulna and the tibia, to the femoral and humeral heads and to the glenoid fossa, and acetabulum.^^ Of particular interest, was the description of a mushroom appearance to the femoral head and an unusually short femoral neck, which is similar to the abnormalities described in ACFD affected individuals.^^'^5 Since the report by Haws and McKusick, an IHH mutation has been identified in affected individuals from this family (AsplOOAsn)(unpublished), suggesting that these malformations occurred due to defective IHH, and not another BDAl gene. Similar abnormalities of the radius, tibia, ulna, glenoid fossa, acetabulum and humerus were noted in a young girl with a complex condition including BDAl, dwarfism, ptosis, mixed partial hearing loss, microcephaly, and mental retardation.^^ However, due to the lack of radiographic evidence, it is not possible to assess whether femoral abnormalities occur in other incidents of BDAl, or whether it is isolated to individuals from these families. Similar malformation of the middle phalanges, the vertebral column and the femoral head in individuals with BDAl and ACFD suggest that IHH is critical for normal patterning of these skeletal elements. The observation of dominant and recessive allelic variants within a single gene, which cause distinct malformations or syndromes, is not uncommon. Dominant mutations of the gene

150


GDF5I CDMPl have been identified in both brachydactyly type C and DuPan Syndrome, while recessive mutations of the same gene are known to cause Grebe type chondrodysplasia, and acromesomelic dysplasia, Hunter-Thompson type (reviewed in ref. 13). In many of these cases, the dominant phenotype is milder than the recessive one, and may reflect the presence of wild type protein in the dominant conditions. A similar relationship between allelic dominant and recessive conditions exists between BDAl and ACFD. While the BDAl phenotype is largely restricted to the middle phalanges, with some exceptions, the ACFD phenotype affects a broader range of skeletal elements. In addition to short middle phalanges, individuals diagnosed with ACFD exhibit short distal phalanges, microcephaly, and a narrow thorax.^^ Short stature is also more pronounced in individuals with ACFD, than those with BDAl.^^ Although these differences in phenotype may be partially due to inheritance of dominant versus recessive mutations, Hellemans et al predicted that they may also occur because of differences in location of the respective mutations. ^^ They showed that while the mutations causing BDAl occur clustered together in a single pocket along the IHH protein surface, the mutations identified in individuals with ACFD are predicted to affect amino acids on the opposite face of the protein (refer to Fig. 1).^^ Due to the lack of in vitro or in vivo data, it is difficult to determine what effect these mutations have on normal IHIi function. The mild phenotype observed in some carriers of the ACFD mutations, and lack of phenotype observed in others, suggests that either the mutant protein retains some activity, or that haploinsufficiency of the IHH protein is not sufficient to greatly affect bone developmental processes. In contrast, the /////mutations identified in individuals with BDAl are inherited in an autosomal dominant pattern, indicating that the wild type IHH gene cannot complement for these mutations. Gao et al predicted that BDAl mutations may disrupt normal endochondral ossification in the digits through haploinsufficiency of the wild-type IHH protein.^ However, if the recessive mutations identified in ACFD carriers also cause haploinsufficiency, but do not cause an obvious phenotype, than it is more likely that the dominant BDAl mutations disrupt IHH signalling through a dominant-negative effect.

IHH Animal Model Although an animal model with specific BDAl or ACFD mutations has not been described, analysis of the Ihh mouse can provide important insights into how the mutant protein may contribute to the physiology of these two malformations. An Ihh mouse was generated by St-Jacques, Hammerschmidt and McMahon by replacing the first exon of the gene with a neomycin resistance cassette. The mutation was lethal and half of the Ihh'' embryos died between 10.5 to 12.5 days post-coitum (dpc), likely due to circulatory abnormalities. Most of the remaining Ihh'' embryos survived to term but died shortly after birth because of respiratory failure that may have resulted from shortening of the ribs. Those embryos that survived to birth were substantially smaller than their wildtype littermates and showed pronounced shortness of their snout, tail, and limbs. Rounded skulls, misshapen endochondral bones and improper joint formation were also observed. Unfortunately, because these mice lacked proper segmentation of the digital bones, it is not possible to compare the brachydactylous phenotype of individuals with either BDAl or ACFD to the Ihh''' mice. Several similarities between the Ihh'' mice and ACFD-affected individuals were apparent, however, and include dwarfism of the long bones, a narrow thorax, and misshapen bones. St-Jacques, Hammerschmidt and McMahon showed that Ihh'' mice had reduced amounts of chondrocyte proliferation compared to wildtype littermates. They also showed delayed chondrocyte differentiation, disorganization of cells in the growth plate, and a shift of hypertrophic cells to the apical ends. Similar growth plate abnormalities in BDAl- and ACFD-affected individuals may in part account for the short bones observed in each condition. Reduced chondrocyte proliferation and displacement of the hypertrophic zone could stunt growth of the skeletal element by diminishing cell numbers within the growth plate. Disorganization of cells throughout the growth plate could also contribute to the development of short bones by altering the

The Genetics ofIndian Hedgehog

151

ability of developing skeletal elements to expand lengthwise through cell expansion during hypertrophy. The observation of irregular, "cone-shaped" epiphyses and premature epiphyseal closure in individuals with both BDAl and ACFD '^^' '^ '^^ supports the hypothesis that IHH mutations in both conditions cause abnormalities of growth plate organization and proliferation. Fitch further hypothesized that, in individuals with BDAl, shortness may be more pronounced in the middle phalanges than in other bones since these bones are the last to ossify and may be more susceptible to the affects of shortened cartilage analgens.^^ One of the primary differences observed between Ihh mice and individuals with IHH BDAl or ACFD mutations relates to the synthesis of bone matrix. While null mice lacked osteoblasts and subsequently did not produce a bone collar, ' similar observations in individuals with either ACFD or BDAl have not been reported. Moreover, BDAl and ACFD mutations do not appear to affect pancreatic development or circulation in humans despite reports of an annulus encircling the duodenum caused by an extension of the ventral pancreas in 42% o£Ihh''' embryos,^ and circulatory abnormalities in Ihh embryos that die between 10.5 and 12.5 dpc. The lack of phenotype in BDAl and ACFD beyond those observed in the skeleton suggests that either the IHH mutants identified thus far retain some function that allows normal activity in other tissues or there are redundant pathways in humans that do not occur in mice. Further experimentation is required to address this issue.

Acknowledgements Funding for this work was provided by the Canadian Institutes of Health Research.

References 1. Vortkamp A, Lee K, Lanske B et al. Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-reiated protein. Science 1996; 273:613-622. 2. Pathi S, Pagan-Westphal S, Baker D P et al. Comparative biological responses to human Sonic, Indian, and Desert hedgehog. Mech Dev 2 0 0 1 ; 106:107-117. 3. Hebrok M , Kim SK, St Jacques B et al. Regulation of pancreas development by hedgehog signaling. Development 2000; 127:4905-4913. 4. St-Jacques B, Hammerschmidt M , M c M a h o n AP. Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev 1999; 13:2072-2086. 5. C h u n g U I , Schipani E, M c M a h o n AP et al. Indian hedgehog couples chondrogenesis to osteogenesis in endochondral bone development. J Clin Invest 2 0 0 1 ; 107:295-304. 6. Dyer MA, Farrington SM, M o h n D et al. Indian hedgehog activates hematopoiesis and vasculogenesis and can respecify prospective neurectodermal cell fate in the mouse embryo. Development 2 0 0 1 ; 128:1717-1730. 7. van den Brink GR, Bleuming SA, Hardwick J C et al. Indian Hedgehog is an antagonist of W n t signaling in colonic epithelial cell differentiation. N a t Genet 2004; 36:277-282. 8. Gao B, G u o J, She C et al. Mutations in I H H , encoding Indian hedgehog, cause brachydactyly type A - 1 . N a t Genet 2 0 0 1 ; 28:386-388. 9. McCready M E , Sweeney E, Fryer AE et al. A novel mutation in the I H H gene causes brachydactyly type A l : A 95-year-old mystery resolved. H u m Genet 2002; 111:368-375. 10. Giordano N , Gennari L, Bruttini M et al. Mild brachydactyly type A l maps to chromosome 2q35-q36 and is caused by a novel I H H mutation in a three generation family. J Med Genet 2003; 40:132-135. 11. Kirkpatrick TJ, Au KS, Mastrobattista J M et al. Identification of a mutation in the Indian Hedgehog ( I H H ) gene causing brachydactyly type A l and evidence for a third locus. J Med Genet 2003; 40:42-44. 12. Hellemans J, Coucke PJ, Giedion A et al. Homozygous mutations in I H H cause acrocapitofemoral dysplasia, an autosomal recessive disorder with cone-shaped epiphyses in hands and hips. A m J H u m Genet 2003; 72:1040-1046. 13. Shum L, Coleman C M , Hatakeyama Y et al. Morphogenesis and dysmorphogenesis of the appendicular skeleton. Birth Defects Res Part C Embryo Today 2003; 69:102-122. 14. Mundlos S, Olsen BR. Heritable diseases of the skeleton. Part I: Molecular insights into skeletal development-transcription factors and signaling pathways. FASEB J 1997; 11:125-132.

152

Hedgehog-Gli

Signaling in Human

Disease

15. Karp SJ, Schipani E, St-Jacques B et al. Indian hedgehog coordinates endochondral bone growth and morphogenesis via parathyroid hormone related-protein-dependent and -independent pathways. Development 2000; 127:543-548. 16. Bell J. O n brachydactyly and symphalangism, Cambridge, London: University Press, 1951:1-31. 17. Fitch N . Classification and identification of inherited brachydactylies. J Med Genet 1979; 16:36-44. 18. Drinkwater H . A second brachydactylous family. J Genet 1915; 4:323-339. 19. Drinkwater H. Account of a family showing minor brachydactyly. J Genet 1912; 2:21-40. 20. Slavotinek A, D o n n a i D . A boy with severe manifestations of type A l brachydactyly. Clin Dysmorphol 1998; 7:21-27. 2 1 . Tsukahara M , Azuno Y, Kajii T . Type A l brachydactyly, dwarfism, ptosis, mixed partial hearing loss, microcephaly, and mental retardation. Am J Med Genet 1989; 33:7-9. 22. Drinkwater H . An account of a brachydactylous family. Proc Roy Soc Edin 1908; 28:35-57. 2 3 . Hoefnagel D , Gerald PS. Hereditary brachydactyly. Ann H u m Genet 1966; 29:377-382. 24. Armour C M , Bulman D E , Hunter AG. Clinical and radiological assessment of a family with mild brachydactyly type A l : T h e usefulness of metacarpophalangeal profiles. J Med Genet 2000; 37:292-296. 25. Yang X, She C, Guo J et al. A locus for brachydactyly type A-1 maps to chromosome 2q35-q36. Am J H u m Genet 2000; 66:892-903. 26. Grange DK, Balfour IC, Chen SC et al. Familial syndrome of progressive arterial occlusive disease consistent with fibromuscular dysplasia, hypertension, congenital cardiac defects, bone fragility, brachysyndactyly, and learning disabilities. Am J Med Genet 1998; 75:469-480. 27. Sillence D O . Brachydactyly, distal symphalangism, scoliosis, tall stature, and club feet: A new syndrome. J Med Genet 1978; 15:208-211. 28. Haws DV, McKusick VA. Farabee's brachydactylous kindred revisited. Bull Johns Hopkins Hosp 1963; 113:20-30. 29. Raff ML, Leppig KA, Rutledge J C et al. Brachydactyly type A l with abnormal menisci and scoliosis in three generations. Clin Dysmorphol 1998; 7:29-34. 30. Farabee W C . Harvard University, 1903. 3 1 . Hall T M , Porter JA, Beachy PA et al. A potential catalytic site revealed by the 1.7-A crystal structure of the amino-terminal signalling domain of Sonic hedgehog. Nature 1995; 378:212-216. 32. Pepinsky RB, Rayhorn P, Day ES et al. Mapping sonic hedgehog-receptor interactions by steric interference. J Biol Chem 2000; 275:10995-11001. 3 3 . Fuse N , Maiti T , Wang B et al. Sonic hedgehog protein signals not as a hydrolytic enzyme but as an apparent Hgand for patched. Proc Natl Acad Sci USA 1999; 96:10992-10999. 34. Armour C M , McCready M E , Baig A et al. A novel locus for brachydactyly type A l on chromosome 5 p l 3 . 3 - p l 3 . 2 . J Med Genet 2002; 39:186-188. 35. Mortier GR, Kramer PP, Giedion A et al. Acrocapitofemoral dysplasia: An autosomal recessive skeletal dysplasia with cone shaped epiphyses in the hands and hips. J Med Genet 2003; 40:201-207. 36. Williams KP, Rayhorn P, Chi-Rosso G et al. Functional antagonists of sonic hedgehog reveal the importance of the N terminus for activity. J Cell Sci 1999; 112:4405-4414. 37. Zeng X, Goetz JA, Suber LM et al. A freely diffusible form of Sonic hedgehog mediates long-range signalling. Nature 2 0 0 1 ; 411:716-720.

CHAPTER 13

Sonic Hedgehog Signaling in Craniofacial Development Dwight Cordero, Minal Tapadia and Jill A. Helms* Introduction

F

acial malformations comprise a significant proportion of the anomalies observed in children born with birth defects. Although congenital facial disfigurements are relatively common and have been exhaustively categorized, science is just now beginning to unravel the genetic and environmental etiologies as well as the molecular mechanisms underlying the clinical manifestations. The Sonic Hedgehog (SHH) pathway is one important signaling pathway involved in craniofacial development. The aim of this chapter is to provide a concise review of craniofacial development within the context of recent advances in our knowledge regarding SHH signaling in facial morphogenesis. We will discuss what is known about the functions of SHH in normal and abnormal craniofacial development in humans and how nonhuman model systems are being used to build a foundation in which to understand the mechanisms underlying craniofacial malformations in humans.

The Face in Myth and Reality In Homer's Odyssey, the Cyclops Polythemus is depicted as a terrifying, mysterious, one eyed man-eating monster (Fig. 1).^ The tale was intended, like many mythologies, to explain phenomena which were not clearly understood at the time but which certainly existed in human and animal populations (Fig. 2). Centuries later, we are beginning to elucidate the scientific mysteries that underlie the physical malformations that gave rise to the myth of the Cyclops and that are still observed today. We now know that cyclopia and other less severe craniofacial disorders can result from disruptions in the SHH pathway caused by genetic or environmental etiologies, or a combination of these factors. Cyclopia is the most extreme manifestation of the congenital disorder holoprosencephaly (HPE), which occurs when the forebrain fails to separate into two distinct cerebral hemispheres. One clinical aspect of this disorder is the wide spectrum of phenotypes involving midline patterning defects observed in fetuses and newborns (see Fig. 3). The spectrum of facial phenotypes ranges from cyclopia with a proboscis, to lesser degrees of facial dysmorphologies consisting of microopthalmia (small eyes), coloboma (congenital abnormality of the eye caused by a lack of complete optic cup closure), hypotelorism (eyes positioned close together in comparison to standard measurements), midface hypoplasia, cleft lip and palate (Fig. 3A,B), and a single incisor."^'^ At the other end of the spectrum, individuals with HPE can have a normal or nearly normal appearing face (Fig. 3C,D). In this chapter we use HPE as a prototype in which to think about the molecular mechanisms governing normal and abnormal craniofacial development. *Corresponding Author: Jill A. Helms—Stanford University, 257 Campus Drive, Stanford, California 94305, U.S.A. Email: [email protected].

Hedgehog'Gli Signaling in Human Disease, edited by Ariel Ruiz i Altaba. ©2006 Eurekah.com and Springer Science+Business Media.

154

Hedgehog'Gli Signaling in Human Disease

Figure 1. The Cyclops Polythemus. The Cyclops Polythemus was depicted in the ancient texts as being a monster with horrific facial features. Several centuries later, modern science is unraveling the molecular and cellular basis for a number of human craniofacial disorders, such as Holoprosencephaly, which may have given rise to the myth of the Cyclops. Reproduced, with permission, from reference 3a.

Figure 2. Facial duplications. Facial duplication has also held a special fascination and involves humans and other animals. Duplications shown in: A,B) human; C) pig; D) cow; E) chick.

Sonic Hedgehog Signaling in Craniofacial Development

155

Figure 3. Spectrum of facial phenotypes associated with holoprosencephaly (HPE). The craniofacial phenotypes observed in patients with Holoprosencephaly (HPE) (A-D) do not always correlate with the underlying forebrain abnormality and can be recapitulated using the chick model system (E-G). A) Children with alobar HPE (the most severe forebrain abnormality in HPE) can manifest with facial malformations including microcephaly, midface hypoplasia, hypotelorism and midline cleft lip and palate as seen in this child. Patients with alobar HPE may exhibit more severe facial malformations such as cyclopia (not shown). B) This child has semilobar HPE (the intermediate form of forebrain malformation in HPE) yet exhibits facial malformations similar to the child in (A). C) A child with lobar HPE (the least severe forebrain malformation in HPE) presents with relatively normal facial features. D) The variation in the clinical phenotype among patients with the same anatomic classification of HPE is seen in this child who also has lobar HPE, but compared to (C) has mild hypotelorism and midface hypoplasia and abnormal dentition. E,F) Stage 25 chick embryos treated with cyclopamine at stage 15 exhibit features reminiscent oflobar HPE, including a single telencephalic vesicle, midface hypoplasia, and hypotelorism. G) Stage 25 chick embryos treated with cyclopamine at stage 17 exhibit features reminiscent of semilobar HPE, including microcephaly, hypotelorism and mild midface hypoplasia. A) courtesy Dr. Alan Shanske; B-D) courtesy Dr. Jin Hahn; E-G) reproduced, with permission, from reference 73. H P E can be classified based o n anatomic or genetic criteria. Either classification scheme is valid within the given context of the patient but the classification criteria should be noted when discussing H P E and the various aspects of the disorder. For example, in the classic or anatomic classification of H P E , the degree to which the telencephalon (forebrain) fails to separate into individual hemispheres has been used as the criteria to define alobar, semilobar, or lobar H P E . For many years, the severity of the facial malformations in H P E was thought to directly correlate with the severity of these brain malformations. This viewpoint led to the postulate, "the face predicts the brain".^ W i t h advanced imaging technologies we now realize that the face does not always predict the anatomic status of the brain, nor does the brain always predict the severity of facial manifestations in H P E . In the genetic classification, the type of H P E is defined based on cytogenetic deletions or mutations in genes known to cause H P E such as SHH^ ZIC2y PTCH, TGIF, or SIX3. For example, mutations in SHHSLTC designated as H P E 3 and may manifest with severe anatomic anomalies of the forebrain and face while in other cases, the same mutation can residt in lesser degrees of brain malformations and a spectrum of facial dysmorphologies which d o not necessarily meet the anatomic criteria for H P E . O n e is left with an incomplete understanding, therefore, of how mutations in SHH can generate such a remarkable diversity of craniofacial phenotypes. In the next section we will discuss how the ability to detect genetic alterations in the S H H pathway raises several new questions about the roles of S H H signaling in brain and facial development.

156


The Craniofacial Consequences of Disrupting SHH Signaling: The Human Experience Genetic Disruptions A number of mutations in the SHH pathway that lead to HPE"^'^'^^ and associated facial phenotypes,^^'^^ have been identified. It has been proposed that the severity of HPE phenotypes is dependent on the extent to which the function of the proteins in the SHH pathway are altered. TraifFort and her group recendy addressed the functional aspects of HPE-causing SHH mutations in an effort to understand their functional consequences.^^ She found that SHH mutations in patient's with HPE fell into one of three classes: those mutations that influenced zinc binding of the protein, those that affected auto-processing of SHH, and those that adversely altered SHH stability. ^^ Her study confirmed what scientists and clinicians had come to accept:'^ there is no clear genotype-phenotype correlation in HPE. This finding strongly suggests that additional genetic and/or environmental factors are important determinants of a given HPE phenotype. In addition to SHH, mutations in other genetic components of the SHH pathway can result in HPE. For example, perturbations in Patched {PTCHl)^ which encodes the receptor for SHH, and the zinc finger transcription factors GLI2 and GLI3 are known to cause HPE and other craniofacial dysmorphologies. Mutations that render the PTCHl receptor unable to bind SHH^^ or adversely alter intracellular interactions of PTCHl may lead to the repression of SMOOTHENED (SMO) and inhibition of SHH signaling, resulting in HPE.^^ To date, no mutations in GLH have been identified as causative for HPE, however, patients with mutations in GLI2 exhibit features of HPE and pituitary anomalies. ^^ HPE is not the only disorder that results from aberrations in the SHH pathway. Mutations in Ptch which lead to truncations of the C-terminus render the protein unable to repress Smoothened activity and the net result is constitutive activation of the SHH pathway. Basal cell nevus syndrome (also known as Gorlins' syndrome) is an autosomal dominant disorder caused by such a gain-of function mutation in PTCHlP'^ The most serious complication of this disorder is the predisposition to basal cell carcinomas and medulloblastomas. These patients may also exhibit dysmorphic facial features consisting of strabismus, cleft palate, and keratocysts ofthejaw.''-^" Mutations in GLI3 also cause a number of craniofacial syndromes. Two disorders are Greig cephalopolysyndactyly syndrome (GCPS) and Pallister-Hall syndrome (PHS), both of which have distinctive craniofacial dysmorphologies.*^^'^^ GCPS syndrome is an autosomal dominant disorder characterized by macrocephaly, a broad nasal root, hypertelorism, a prominent forehead, pre-and postaxial Polydactyly, and syndactyly of the hands and feet."^ '"^^ The syndrome shares some clinical features with Pallister-Hall syndrome (PHS) such as craniofacial abnormalities, postaxial polysyndactyly and inheritance in an autosomal manj^g^ 22,26 f^Qêver, PHS is clinically distinguishable from GCPS in that PHS patients may have a hypothalamic hamartoblastoma (tumor), do not exhibit hypertelorism or a broad nasal root,"^^ and occasionally have midline cleft lip and palate holoprosencephaly. Studies suggest that GCPS syndrome results from functional haploinsufficiency o^GLI3^^ '^^ while PHS results from the production of a truncated GLI3 protein. Truncation mutations can, however, cause GCPS. Based on GLI3 truncation mutations and the clinical phenotypes of these disorders, a model which showed that mutations in GLI3 can mimic the bi-functional nature oi Drosophila Cubitus interuptus (Ci) was generated. ^^ It was found that full length GLI3 localizes to the cytoplasm and activates PTCHl expression, whereas the mutant GCPS-GLI3 protein had no effect on PTCHl expression, consistent with haploinsufficiency in this disorder. However, the mutant PHS-GLI3 protein repressed GLI3 activated PTCHl expression."^^ Smith-Lemli-Opitz Syndrome (SLO) is an autosomal recessive disorder, which may be associated with perturbations in SHH signaling. HPE occurs in approximately 5% of SLO


157

patients.^^"^^ SLO is due to a defect in 3 beta-hydroxy-steroid-A7-reductase (7-DHC reductase) that leads to impaired cholesterol synthesis.^'^'^^ Cholesterol plays many roles during embryonic development, and most importantly for this discussion, in SHH signaling. Cholesterol appears to play a role in the cellular transport of Shh and in the generation of or in allowing clustering of the morphogen to create areas of higher and lower concentration in a cell. Although the relationships between cholesterol and SHH are not completely characterized, perturbations in cholesterol metabolism may alter SHH function.^^'

Teratogens and Their Effect on SHH Signaling A number of teratogens have been identified which appear to exert their adverse effects on craniofacial development via the SHH signaling pathway. The untoward consequences of maternal teratogen exposure depend upon three factors: first, the genetic susceptibility of a given embryo to a teratogen; second, the developmental stage of the embryo at the time of exposure; and third, the mechanism of action of the teratogen. While birth defects are caused by inadvertent exposure to teratogens, these substances are being increasingly exploited in the laboratory setting to gain insights into both normal and abnormal craniofacial developmental processes. Alcohol is a known teratogen, yet prenatal exposure still occurs at a daunting rate, making it one of the most common causes of preventable birth defects and mental retardation. Fetal Alcohol Syndrome (FAS) occurs between 0.2-1.5 per 1000 live births and is characterized by growth restriction, abnormalities of the central nervous system, and a stereotypical facial dysmorphology. The craniofacial features observed are microcephaly, microopthalmos, hypoplastic philtrum and maxillary hypoplasia. The underlying molecular mechanisms resulting in the clinical manifestations of FAS remain enigmatic, although recent studies are providing more clues. Exposure to ethanol alters the embryonic expression patterns of Shh in the craniofacial complex. ^ The net effect of altering Shh expression appears to be the premature death of cranial neural crest cells. Although these studies were conducted in nonhuman models they may provide a molecular or cellular explanation for how alcohol adversely affects human brain and facial development and will be discussed in detail later in this chapter. Malformations of the craniofacial region are also seen in cases of excess and deficient vitamin A, and the phenotypes of the effected individuals appear to be remarkably similar to individuals with FAS. ^' Public attention to the teratogenic potential of vitamin A and its derivatives came in the early 1980s when women became pregnant while taking the drug Accutane (Isotretionoin) for treatment of acne. In addition to increased spontaneous abortions, children born to mothers who took Accutane had birth defects involving the brain, heart and face. Exogenous retinoic acid, the active metabolite of vitamin A, has been shown to disrupt the expression of Shh in chicken embryos and may play a role in the clinical manifestations seen in either excessive or deficiencies in vitamin A or its derivatives. A group of drugs used to treat hypercholesterolemia, the Statins, pharmacologically lower cholesterol biosynthesis by inhibiting HMG-CoA reductase, and may be associated with limb and central nervous system (CNS) malformations such as HPE when taken during the first trimester of pregnancy.^ Presumably the teratogenesis is due to decreased cholesterol bioavailability needed for a number of processes during embryogenesis. Theoretically this may affect autoprocessing or transport of Hedgehog (Hh) proteins thus adversely influencing Hh activity. The teratogenic potential of these drugs in humans requires further clinical and basic scientific investigation. Taken together, these clinical findings indicate a strong correlation between disruptions in SHH and craniofacial malformations. These studies also implicate a wide variety of drugs that can have a teratogenic effect, and do so presumably by altering SHH signaling. In order to rigorously test our understanding of how SHH functions in craniofacial development, however, model systems have had to be developed which allow us to alter spatially and temporally SHH signaling. Some of these models are presented in the following section.

158


Figure 4. Targeted disruption in genes in the Sonic Hedgehog (Shh) pathway results in craniofacial dysmorphologies in the mouse. A) Wild type mouse embryo at stage 15.5 reveals normal craniofacial anatomy. B) Stage 15.5 Shh null embryo with grossly apparent abnormality of the forebrain, anopthalmia, proboscis, and hypoplasia of the maxillary and mandibular prominences. The fetus is severely growth restricted and has distal limb anomalies. C) Stage 14.5 embryo with Ptchl null background partially rescued by a transgene expressing Ptchl. The resultant alteration in Shh activity leads to abnormal frontonasal process derivatives such as the nose, hyperplastic maxillary processes and bilateral clefting of the lip and palate. The mandible does not appear to be afFeaed by the underlying genetic disruptions.

Models to Study Craniofacial Development As in other biological disciplines, deducing the constituents of a biological pathway and understanding their functions is the first step in understanding normal as well as abnormal processes during embryonic development. Because it is not always possible to use human subjects, the use of model systems such as chickens (Figs. 3E-G, 5A-F, 6), mice (Fig. 4), fruit flies, worms, and fish are invaluable. Although the models in these systems exhibit widely disparate facial appearances, accumulating evidence suggests that most genes involved in craniofacial development are evolutionary conserved across species. Therefore, despite differences between mouse and man, or birds and fish, this remarkable amount of conservation allows us to unravel the mechanisms underlying normal and abnormal craniofacial development using a number of species. Consequently, a considerable amount of the knowledge about human craniofacial development is attributable to the extensive study of developmental processes in such model systems. We will discuss recent experimental data from these model systems, which has lead to our understanding of both normal and abnormal craniofacial morphogenesis.

Craniofacial Consequences of Shh Perturbation: The Model Experience Genetic Disruptions A number of mice have been generated in which mutations in the Shh pathway result in HPE '^ and other craniofacial malformations (see Fig. 4). Targeted disruption of the Shh gene in mice (Fig. 4B) has revealed that Shh plays a number of critical roles in brain, axial skeleton, limb and craniofacial development. Null mutations in Shh lead to malformations of the brain, vertebral column, limbs as well as cyclopia and malformations involving the first pharyngeal arch derivatives (Fig. 4B). As in humans, mutations in molecules downstream of the Shh ligand also residt in malformations (see Fig. 4C). For example, mutations in the murine Ptchl gene result in open and overgrown neural tubes. Mice homozygous for the Ptchl mutation exhibit ventralization of the neural tube and the ventral forebrain, with dorsal expansion of the ventral forebrain marker My2.i.^Ânumberof mice heterozygous for the Ptchl mutation develop tumors of the cerebellum known as medulloblastomas. These studies reveal the importance of the interplay between Shh and Ptch in regulating growth and patterning during neural plate and neural tube


159

development. ^ However, the ability to study the roles of these molecules during later craniofacial development is limited because of the disruption of early neural patterning in Shh mutant mice and loss of viability past E9.5 in mice homozygous for PtchL To address the role of Hedgehog (Hh) signaling on the cranial neural crest cells (CNCCs) during craniofacial development without disrupting early neural patterning, the ability of the CNCCs to respond to Hh signaling was removed by generating Wnt-1-Cre; conditionally null Smoothened mice (Wnt-1-Cre; Smo^^").^^ Embryos carrying the conditional knockout of smoothened exhibited extensive loss of craniofacial structures. This experimental strategy demonstrated the requirement of Hh signaling in postmigratory CNCCs and the roles of Hh signaling on craniofacial development during later stages of embryogenesis.^^ These authors also show that Fox genes play a role in mediating Hh activity during craniofacial development. To circumvent embryonic lethality in Ptchl niJl mice, Scott et al partially rescued the phenotype by overexpressing P^r^7 using a metallothionein promoter to drive Ptchl expression in a Ptch-null background. They utilized two lines of mice that differed in their copy number (i.e., high and low) of the Ptchl transgene, and were able to obtain viability up to stage E14-14.5.^^ Use of the line with a lower copy number, and therefore less Ptchl expression/ rescue, lead to an increase in viability to El2.5 but the embryos still had neural tube defects (NTDs). Embryos with the higher copy number survived to E14.5 without NTDs.^^ Further examination of these high copy number transgenic mice,^^ found evidence of maxillary hyperplasia reminiscent of the large maxillary processes in Gli3-^- mutant mice (Fig. 4C). This observation suggests that growth of the maxillary processes, from which the sides of the face and the secondary palate are derived, are finely regulated by the activity of the Shh pathway. We also noted that the mandibular process, which gives rise to the lower jaw, was unaffected by Ptchl disruption (Fig. 4C).^^ These findings suggest that the factors controlling growth in the different facial processes may be unique and separable. Conversely, overexpression of Ptchl in a Ptchl wild type background causes inhibition of the induction of Shh target genes, which manifests as a dorsalization of the neural tube along with a failure of tube closure around the telencepahlon.^^ Shh signaling has also been altered in mouse models by mutations in Gli3.^^ Mice carrying a Gli3 null mutation are reminiscent of human GCPS.^ As such, insight into the molecular etiologies underlying the GCPS phenotype has been aided by studying these "Extra-toes" mice. In these Gli3 null embryos F ^ ^ domains are upregulated in the anterior neural ridge and facial primordia, and these regions display reduced apoptosis. Conversely, FgfB expression is decreased in Shh'^' mice,^^ which suggests that in addition to its ability to limit the activity of Shh, Gli3 may also restrict the expression of/^^. In Gli3 null mice, Pax9 expression is expanded in the medial nasal process, consistent with the hypertelorism phenotypes of GCPS syndrome patients.^^ Mutations in Gli2 result in failure of the floor plate to develop. Gli2 mutant mice also show a decrease in Shh signaling, as manifested by a ventral shift in expression o{ Nkx2.2 and Pax6m the neural tube.^ This suggests that Gli2 protein fimctions as a transcription activator to regulate the expression of Shh-responsive genes.^ There are no data currently available on the ftmction of Gli2 in the craniofacial complex, separate from its earlier role in neural tube patterning. Abnormal Shh signaling is also observed in talpid3 chicken mutants. ' These mice exhibit Polydactyly and abnormal facial morphology consisting of fused nasal placodes and hypotelorism as well as hypoplastic maxillary primordia which are fused in the midline.

Teratogenic

Disruptions

For several years, teratogens have been used in a number of model systems in order to study developmental mechanisms and to understand how teratogens exert their adverse effects. In this section we present three known environmental teratogens and one compound which is not currendy recognized as such, and describe how they alter Shh function. A short discussion of

160


Figure 5. Sonic Hedgehog (Shh) signaling in the craniofacial complex is required for normal patterning of the upper beak. A,C,E) are oblique views of stage 41 chick embryos and, B,D,F) are corresponding skeletal structures of these embryos visualized with alcian blue and alizarin staining. A,B) Control embryos reveal the normal gross morphology and skeletal structures. C,D) The morphologic consequences of exposing stage 17 embryos to cyclopamine consist of mild microcephaly and hypotelorism as well as malpositioning and truncation of the distal aspect of the upper beak. Proximal elements such as the dorsal component of the premaxillary bone (e.g., the nasal process of the premaxilla (pn)) is intact, while the body of the premaxillary bone (pm) is shifted ventrally. The mandible (mn) is unaltered in cyclopamine-treated embryos. E,F) Following placement of 5E1 cells, which produce antibodies to Shh at stage 9.5, microcephaly, abnormal eye development and abnormalities in the distal upper beak are seen (E). The gross phenotype is consistent vdth the morphological changes seen following inhibition of Shh signal transduction with cyclopamine (compare to (C,D)). A-D) Reproduced, with permission, from reference 73; E-F) Reproduced, with permission, from reference 106. their use and implications is presented here; in the case of cyclopamine, more detail will be provided in a separate section. The first teratogen we will discuss is perhaps the best known: alcohol. Prenatal alcohol consumption in humans leads to FAS and lifelong physical and mental impairment. Although education about drinking in pregnancy is ongoing and abstinence is preventive, pregnant women continue to drink and their children continue to bear the consequences. Model systems to study FAS are uncovering the mechanisms underlying these impairments and may involve alterations in Shh signaling. Administration of ethanol to chicken embryos resulted in down-regulation oi Shhy Ptchy GUI, Gli2, and Gli3 expression in the developing head, neural crest cell death and craniofacial defects. ^ Shh overexpression or administration of recombinant Shh can rescue neural crest cell death and in general, lessen the craniofacial manifestions of the alcohol exposure. Alcohol may also exert effects by interfering with retinoic acid (RA) metabolism, in eflPect usurping the function of retinol dehydrogenase and causing decreased RA in the embryo.^^ The consequences of retinoid disruption have been studied in some detail in the face. For example, in the chicken teratogenic doses of RA inhibit Shh expression and abolish polarizing activity in the frontonasal and maxillary processes. Inhibiting endogenous retinoic acid receptor activity in the heads of chicken embryos leads to loss of Shh expression in the facial ectoderm. The morphological consequences of this biochemical disruption are decreased outgrowth of the mid-and-upper facial structures. In zebrafish, exogenous retinoic acid administration results in an initial decrease in Shh, but later an ectopic expression of the morphogen. Taken together, these data suggest that RA may exert its teratogenic effects through


161

the Shh signaling pathway and that the response of Shh to RA may be biphasic in some model systems. Another potent teratogen is cyclopamine (11-deoxojervine), a steroidal alkaloid extracted from the plant Veratrum californicum. As its name suggests, q^clopamine is capable of inducing cyclopia when embryos are exposed during early development (see Figs. 3E-G, 5C,D). Cyclopamine first came to attention in the 1950s when it was observed that pregnant ewes grazing on Veratrum californicum on day 14 of their gestation gave birth to sheep with cyclopia. This led to the eventual purification of cyclopamine and other steroidal alkaloids responsible for this defect. Since that time, cyclopamine has been a powerful tool to study the effects of altering Shh signaling in the brain and face in a number of model systems (avian, murine, amphibian, and fish) (see Figs. 3E-G, 5CyD).^^'^^^^ Cyclopamine inhibits Shh signal transduction by binding directly to Smo and possibly altering its protein confirmation. The efficacy of steroidal alkaloids to inhibit Shh transduction and phenocopy mutations in Shh was shown by exposing gastrulation-stage chick embryos to another steroidal alkaloid, jervine, that recapitulated the anomalies seen in the Shh'' mouse. In the chicken model, temporal interruption of Shh with cyclopamine reproduced the spectrum of HPE-associated phenotypes seen in humans (Fig. 3E-G).^ The ability of these compounds to block Shh may also have clinical implications in human diseases because some cancers are associated with uncontrolled Shh signaling. Drugs that interfere with cholesterol metabolism are the third class of potential teratogens. Cholesterol is important in many aspects of embryonic development,^ and is involved in autoprocessing and transport of Shh.^^ The compounds AY9944 and BMl5.766 block the function of A7-reductase, which itself is required for cholesterol synthesis. Rats exposed to AY9944^^ or BMl5.766^^ produce the same biochemical disruptions as seen in SLO and may lead to HPE.^^ The low-density lipoprotein receptor family member Megalin (gp330) is important in embryonic development. ^^'^^ Megalin is expressed in the visceral endoderm, neural ectoderm, neural plate and neural tube.^'^ It mediates endocytosis of ligands, targeting them for lysosomal degradation or transcytosis.^^'^^ Megalin has been shown to bind to Shh with high affinity and mediate endocytosis.^^ It also binds to complexes of retinal-binding protein.^ ' The receptor-associated protein (RAP) is a chaperone, which regulates the expression and processing of LDL receptor-related protein and can bind to Megalin. Administration of the receptor-associated protein (RAP) to gastrulation stage chick embryos produces cyclopia (Cordero and Helms, unpublished observations). The teratogenic effects elicited by exogenous RAP appears to result from inhibition of the receptor-ligand interaction between Shh and megalin and or with other complexes and therefore most likely accounts for the resultant phenotype. Megalin may therefore provide a link between Shh signaling and RA. The significance of megalin-Shh interactions is discussed in the final chapter of this book.

Alternative Methods of Disrupting Shh A number of different approaches to disrupting Shh signaling are also being applied experimentally and the reader is referred to the articles below for a detailed description of these techniques. One method that has been used successfully to block Shh activity is the use of anti-Shh antibodies (Fig. 5E,F).^^'^^ RNA interference (RNAi) or gene knock-down is another method to silence a gene of interest.^^"^^

Embryologic Aspects of Normal Craniofacial Development: An Introduction During development of the craniofacial complex, the transformation of single pluripotent neural crest cells into a complex three-dimensional structure such as the face requires an elaborate interchange between tissues that form the face.^^'^ '^^ Shh plays a number of important roles in pattern formation and growth of the face, by mediating interactions within and between these tissues.

162


In general, the fate of neural crest cells in the face is controlled by communication with nearby cells/tissue. The fates of these craniofacial neural crest cells are highly dependent upon the developmental time point of which this fate information is received, and one mechanism by which signaling molecules such as Shh control patterning and growth of the face is through organizing centers and epithelial-mesenchyme interactions.

Organizing Centers During embryonic development, groups of cells have the capacity to induce biochemical and morphological changes in neighboring cells. Based on this function, these groups of cells have been referred to as "organizers" or "organizing centers". Organizers are critical determinants of cell fate and provide positional information that is required for the establishment of body axes. The organizers' ability to alter the fate of other cells was initially shown in 1924 by Spemann and Mangold, who demonstrated that transplanted dorsal blastopore lip of the amphibian induced a secondary developmental axis.^ Since that time, a number of embryonic organizing centers have been identified which exert their developmental influences through Shh. These centers include: the notochord, the zone of polarizing activity in the limb bud, Hensens node, the tail bud, and the frontonasal ectodermal zone. Shh secreted by these organizing centers may establish concentration- or activity-dependent gradients that determine different cell types and positional information. Shh can act direcdv on target cells to produce distinct cellular responses in a concentration-dependent fashion, but the concentration of Shh may not correlate direcdy with the gradient of activity sensed by the target cells. It is possible that the concentration gradient can be established or influenced by the levels of Ptchl ' and growth-arrest specific protein 1 (Gasl), both of which bind to Shh and influence Shh activity.

The organizing properties of Shh have been demonstrated in the neural tube. During neural tube development Shh is initially secreted by the notochord, which subsequendy induces the floor plate cells of the neural tube to secrete Shh.^^^ A concentration gradient of Shh is established within the neural tube, which specifies progenitor cell identity and neuronal fate through a set of homeodomain transcription factors. Shh establishes ventral polarity in the neural tube and in conjunction with bone morphogenic proteins (Bmps) secreted from the roof plate^^^ delineates dorsal-ventral polarity of this structure. Development of the forebrain may involve similar signaling mechanisms. ^^ Organizer centers which are located outside or within the craniofacial complex provide information for patterning and morphogenesis of the face, most likely through the establishment of morphogen gradients similar to those of the notochord and floor. Because the structural anatomy is quite distinct between the neural tube and face, additional mechanisms may influence just how this Shh-mediated organizing activity is established and interpreted. These anatomical distinctions are discussed in more detail in the following paragraphs. The prechordal plate mesendoderm underlying the forebrain is a source of Shh which patterns the forebrain. '^^^ Loss of Shh in the prechordal plate in Shh mice results in severe neural plate alterations which causes brain malformations and cyclopia. In turn, Shh in the ventral forebrain has a significant influence on morphogenesis of the frontonasal primordium, from which the middle and upper face develop.'^^' The endoderm also has head organizing activity and presumably is also a source of Shh signaling crucial to head development. Discrete domains of the craniofacial epithelia exhibit Shh-mediated organizing activity. This was demonstrated by grafi:ing Shh-expressing facial ectoderm into the anterior region of a limb bud, where the tissue exhibited organizing activity as assayed by its ability to induce digit duplications. Precisely how this organizing activity is established is still not clear. We subsequently determined that this facial organizing center is delineated by a boundary between Shh and Fibroblast growthfactor 8 {FgfB) expression, ^^^ and referred to this region as the frontonasal ectodermal zone, or FEZ (Fig. 6A).


163

Figure 6. Transplantation of the facial ectodermal zone (FEZ) demonstrates its importance in patterning of the frontonasal primordia. A) In situ hybridization on a sagittal section of a stage 20 chick embryo; red corresponds to a 5M-expressing domain and green corresponds to a i^^-expressing domain. B,E) are representations of similar lateral sections of stage 20 and stage 25 chick embryos, respectively. The FEZ is indicated by dotted lines, where yellow corresponds to the 5M-expressing domain and green corresponds to the i^^-expressing domain; the junction between Shh and Fgf^ is indicated with an arrowhead. C,F) Photographs of stage 36 control and transplanted embryos, respectively. Arrowhead indicates formation of an ectopic beak from transplanted FEZ. D,G) Trichrome stained sections of stage 36 control and transplanted embryos, respectively. Again, arrowhead indicates formation of an ectopic beak from transplanted FEZ. A-G) Reproduced, with permission, from reference 108.

164


The FEZ plays a critical role in establishing dorsoventral patterning of the frontonasal primordium.^^^ Heterotopic transplantation of the FEZ from stage 20 quail embryos into stage 25 chick embryos provoked a repatterning of first pharyngeal arch mesenchyme which resulted in the generation of an ectopic beak whose dorsal-ventral polarity directly correlated with the orientation of the Shh and Fgj^ expression domains (Fig. 6). ^^

Epithelial'Mesenchytnal Interactions The importance of epithelial-mesenchymal interactions in craniofacial development was shown bv classic recombination studies between the oral epithelium and underlying mesenchyme. As with all epithelial-mesenchymal interactions, they provide the correct spatial and temporal cues necessary to generate normal anatomic structures and their relationships. Shh in conjunction with a number of growth factors including Bmps, Wnts, and Fgfs mediates these epithelial-mesenchymal interactions.^ ^ The importance of Shh signaling during development of the forebrain facial prominences,"^^'^^'^^'^ palate,^ ^^ teeth, '^^^ taste papillae^^^ and submandibular salivary glands are becoming more evident and important in understanding clinical manifestations in several craniofacial disorders.

Forebrain Development and Shh The commencement of the craniofacial region is considered to coincide with gastrulation. At this stage, the embryonic ectoderm is divided into neural and non-neural ectoderm. The neural ectoderm will give rise to the forebrain, whereas the non-neural ectoderm will give rise to the epidermis of the face and the rest of the body. The close spatial proximity of these two ectodermal regions gives rise to an intimate molecular and structural relationship between the developing forebrain and the presumptive facial regions. During development of the central nervous system, the anterior prosencephalon becomes the telencephalon (cerebral hemispheres and basal ganglia) and the diencephalon (thalamus, hypothalamus and pituitary gland).^^^ The prechordal plate mesendoderm is located in the ventral midline of the forebrain beneath the neural plate, and is a source of Shh. Shh from the prechordal plate is critical for establishing ventral polarity during dorsal-ventral patterning of the forebrain and proper development of the dorsal telencephalon. During normal development of the dorsal telencephalon, the roof plate undergoes invagination (commonly referred to as cleavage), which in conjunction with growth of the telencephalon, leads to two separate cerebral hemispheres.^ ^^'^^ Interruption of Shh can result in the failure of invagination of the dorsal telencephalon and a single hemisphere.'^ The mechanism by which Shh influences the dorsal telencephalon is unclear since Shh is expressed ventrally. A number of other roles for Shh during neurogenesis exist but are outside the scope of this chapter.'^ "'^^ In avians, Shh expression is initially restricted to the ventral diencephalon at stage 15 (Fig. 7C,D).^^ At stage 17, a separate domain of Shh is induced in the ventral telencephalon, which is separated from the diencephalic domain by a ^M-negative optic recess (Fig. 7E,F).^^ At stage 20, Shh expression is first detected in the facial ectoderm (Fig. 7G,H). This progressive format to Shh expression in the craniofacial tissues (Fig. 7) underscores how many tissues are responsive to this morphogen.

Craniofacial Development and Shh In mammals, structures of the face develop from the facial prominences (Fig. 8). They include the paired maxillary, the paired lateral nasal, the paired mandibular and a single frontonasal. These primordia consist of undifferentiated cranial neural crest cells or mesenchyme enclosed by epithelium. The mesenchyme is derived from undifferentiated cranial neural crest cells that originate from the dorsal neural tube. These cells migrate into specific regions of the facial primordia and differentiate into the cartilaginous and osseous structures of the face. '^^ '^^^ During their migration cranial neural crest cells are interposed between the neuroectoderm and facial ectoderm until reaching their final destination, either


165

Figure 7. Ontogeny of Sonic Hedgehog {Shh) expression in the chick craniofacial complex. A,C,E,G) are representations of the (B,D,F,H) actual midline sagittal embryonic sections. In figures (B,D,F,H) red represents Shh expression as obtained by in-situ hybridization with S . A,B) At stage 10 Shh is expressed in the ventral prosencephalon (vf), oral ectoderm and pharyngeal endoderm and the prechordal mesoderm (pcm). C,D) By stage 15 the prosencephalon is comprised of the telencephalon (te) and the diencephalon (di). At this stage Shh transcripts are localized to the (di). E,F) At stage 17, Shh is now expressed in the (te). G,H) Around stage 20, Shh is expressed in not just in the diencephalic and telencephalic neural ectoderm, but also in the facial ectoderm (arrow). I,J) Shh expression remains in the neural ectoderm and facial ectoderm at stage 22. Panels B,D) reproduced, with permission, from reference 73.

166


Figure 8. Development of the craniofacial primordia. A-D) Representations of frontal views of mouse embryos showing the prominences that give rise to the main structures of the face. The frontonasal (or median nasal) prominence (pink) gives rise to the forehead (A), the middle of the nose (B), the philtrum of the upper lip (C) and the primary palate (D), while the lateral nasal prominence (blue) forms the sides of the nose (B,D). The maxillomandibular prominences (green) give rise to the lower jaw (specifically from the mandibular prominences), to the sides of the middle and lower face, to the lateral borders of the lips, and to the secondary palate (from the maxillary prominences). E) Frontal view of a chick embryo, also showing which prominences give rise to different facial structures. F) Frontal view of a human child, with different facial structures colored to indicate the prominences from which each structure developed. A-F) Reproduced, with permission, from reference 147. the pharyngeal arches or the frontonasal primordium. Developmental instructions are received from the surrounding epithelium during their passage and foUovîng arrival have a profound influence on their developmental fates. In the avian model system cranial neural crest cells are thought to provide patterning information that directs morphology of regions of the craniofacial complex. ^^^'^ As a consequence of both passive and active cell movements, the forebrain neuroectoderm and the facial ectoderm remain in close proximity, being displaced "in register" during head flexure. T h u s the brain, rather than being a passive scaffold, actually influences facial patterning via a complex dialogue wîth facial tissues. This new^ view is laying the foundation for investigations into hov^ genetics and epigenetics influence the dynamic interactions between the brain and face. W h e n Shh expressing ectoderm in the frontonasal primordium is removed, the result is inadequate outgrowth of the middle and upper face (Fig. 9D-I).^^^ Conversely, supplying exogenous Shh causes an expansion in the mediolateral width of the face (Fig. 9J-L). When over-expression o^Shh is localized to the facial ectoderm, by delivering Shh via a retrovirus, the same morphogenic consequence of a widened frontonasal primordium is observed (Fig. 9J-L).^^^ Similar effects can also be produced by manipulating genes that regulate Shh activity, as seen in Gli3" mice.^ '^^^ But how Shh is established in the face, and how this facial domain coordinates growth and development along with the forebrain domain of Shh has, until recently, remained unclear.


167

Figure 9. Consequences of removing and overexpressing Shh in stage 30-35 chick embryos. A-C) Frontal (A), oblique (B), and ventral (C) views of normal stage 25 chick embryos. D-F) Consequences of removing 5M-expressing frontonasal (fnp) ectoderm at stage 20. Removal of fnp ectoderm causes a loss oi Ptc expression (arrow, D) by stage 30. Excision also causes a visible cleft (arrow, E) between the maxillary process (mxp) and lateral nasal process (Inp) at stage 35, as well as bilateral cleft palate (arrow, F). G-I) Consequences of removal of 5/?^-expressing mxp ectoderm. Removal of mxp ectoderm causes a loss oiPtc expression (arrow, G) by stage 30. Excision also causes a similar cleft (arrow, H) between the mxp and Inp by stage 35, as well as unilateral cleft palate (arrow, I). J-L) Consequences of injecting RCAS-5M fibroblasts into fnp ectoderm. Ectopic administration oiShh results in considerable widening of the upper beak (J) along with growth of an ectopic beak (arrow, K). Hypertelorism (not shown) also results from expansion of the fnp, and the palatal bones are positioned more laterally (arrow, L). A-L) Reproduced, with permission, from reference 131. Shh in mice results in early neural plate patterning defects that result in a single telencephalic vesicle and severe craniofacial malformations (Fig. 4B). T h e early loss of Shh at the neural plate stage in this model precludes delineation of the role(s) of Shh in later facial development.

168


In order to circumvent this limitation imposed by the interruption of the neural plate, other model systems and techniques have been used to interrupt Shh at later stages of embryonic development. The chick was recendy used to study the effects of temporal inhibition of Shh transduction on facial morphogenesis, by administering cyclopamine at specific developmental stages. The stages chosen for cyclopamine treatment were based upon the elucidation of the dynamic expression pattern o^Shh in the brain and face."^^ If Shh was blocked prior to initiation of expression in the telencephalon, brain anomalies as well as facial defects such a severe hypotelorism, microopthalmia, and microcephaly were noted (Figs. 3E,F).'^^ However, when Shh was blocked after its induction in the ventral telencephalon but prior to its expression in the facial ectoderm, no gross phenotypic brain anomalies were observed and the facial malformations were less severe than those treated at earlier stages (Fig. 3G, 5C,D).^^ When Shh was blocked after its induction in facial ectoderm then the embryos exhibited only subde facial anomalies (data not shown).^ Taken together, these data suggest that the brain and face are differentially susceptible to perturbations in the Shh signaling pathway. This is clinically relevant since some patients with mutations in SHH sometimes have abnormal brains and facial features while others have normal-appearing brains and abnormal facial phenotypes. In another set of experiments conducted to look at the role of Shh specifically emanating from the forebrain, hybridoma cells producing antibody (5E1) to Shh were injected into the brain of stage 9.0-9.5 embryos and evaluated at state 4l(Fig.5E,F).i^This treatment blocked Shh emanating from the ventral forebrain. We found that the resultant facial dysmorphologies (Fig. 5E,F) following cyclopamine or 5E1 administration are not secondary to increased neural crest cell apoptosis in the frontonasal primordium, but to a molecular alteration in the FEZ.^ For example, inhibition of Shh in the ventral forebrain neuroectoderm blocked Shh expression in the facial ectoderm.^ This molecular alteration was accompanied by a proximal expansion of the FgfB domain, which interrupted the molecular boundaries of the FEZ. ^^ The morphological consequences of interrupting this organizing center are truncation of the upper beak, ventralization of the premaxillary bone, and decreased expansion of the medial-lateral axes (Fig. 5E,F)."^ The intervening mesenchyme is critical for either the induction or the maintenance of Shh expression in the face, as shown by blocking the ability of neural crest cells to respond to a Hedgehog signal, as shown by Jeong^^ and as discussed previously.

Palatogenesis and Shh The definitive palate in mammals consist of the primary and secondary palate. The primary palate forms from the fusion of the medial nasal prominences. ^^^ The secondary palate gives rise to the majority of the hard and soft palate are formed from the palatal shelves which extend from the maxillary processes. ^^^ These shelves structurally consist of an epithelial surface surrounding neural crest cells or mesenchyme. As a result of mesenchymal cell proliferation the palatal shelves initially grow vertically downward on either side of the tongue. Subsequendy, the vertical palatal shelves elevate to a horizontal position above the tongue. In the final stage of palate development, the now horizontally-oriented, elevated palatal shelves fuse to form the palate. Shh plays a role in palatal development and appears to be required for signaling interactions within this structure. Fibroblast growth factor (Fgf) signaling, specifically FgflO and the Fgf receptor 2b (Fgfr2b) together with Shh, are implicated in the outgrowth of the palatal shelves. ^ "^ Fgfr2b is expressed in the epithelium of the palatal shelf between El 2-14, and at lower levels in the mesenchyme of the nasal side of the palate. FgflO, which functions as a ligand for Fgfr2, is expressed in the adjacent underlying mesenchyme. ^^"^ Null mutations in Fgfi'2^^^zxiAFgfld^^^ result in decreased cell proliferation in the palatal mesenchyme, down regulation of Shh in the palatal epithelium, and clefting secondary to decreased outgrowth of the palatal shelves. In vitro experiments using recombinant FGF 10 show that Fgf proteins can induce Shh expression


169

in the palatal epithelium and can cause increased expression oiPtchl and cell proliferation in adjacent mesenchyme.^^^ An increase in cellular proliferation is also elicited with application of exogenous Shh to wild type mesenchymal palatal explants. Thus in the palate, Shh is a downstream target of Fgfr2/Fgfl0. The clinical implications of these findings may be that alterations in Shh and/or Fgfr2/Fgfl0 signaling are involved in select cases of cleft palate, which are presently described as idiopathic. Surprisingly however, studies of newborns with isolated oral clefts suggest that mutations in SHH are not a common cause of nonsyndromic oral clefts; however, further investigation is warranted.

Odontogenesis and Shh Shh plays a number of roles during tooth development such as the location and generation of the teeth. Though the roles of Shh are slightly different in teeth than in palate, Shh is involved in reciprocal signaling interactions between the epithelia and the mesenchyme. During odontogenesis in the mouse, Shh expression is limited to localized thickenings of oral epithelium that will give rise to teeth. However, Shh activity is absent from the diastema mesenchyme, an edentulous region between developing teeth. The spatial relationship of Shh to the presence or absence of teeth highlights the importance of its spatial activity during odontogenesis. Reciprocal interactions between the oral epithelium and mesenchyme occur during tooth development. Mandibular mesenchyme appears to have the ability to decrease and/or limit Shh signaling via interactions with the epithelium.^^^ When mandibular processes were cultured without the overlying epithelium, Ptchl and GUI were up-regulated, and the expression of Gasl was down-regulated in the underlying diastema mesenchyme.^^^ Taken together, these data suggest that Gasl in the mesenchyme has an antagonistic effects on Shh signaling in the epithelium, a relationship that has been observed in other tissues.^^ Shh appears to initiate tooth bud formation and to influence the morphology of teeth. For example, when Shh signaling is blocked in the epithelium the number of teeth that develop is substantially reduced.^ Blocking Shh activity at slightly later stages allows some teeth to form, but the superficial epithelium of the developing tooth buds often exhibits apoptosis.^^ The role of Shh during tooth morphogenesis is extensively covered in a later chapter.

Papillogenesis and Shh The tongue develops from contributions of the first four pharyngeal arches and involves complex epithelial-mesenchymal interactions. The surface of the tongue has multiple types of papillae, which provide mechanical functions (filiform) and others that contain tastebuds and are known as gustatory papillae (fungiform and circumvallate). Papillary development begins when localized regions of epithelial cells begin to proliferate, giving rise to epithelial thickenings called placodes. The placodes subsequently evaginate into the mesenchyme thereby forming the raised papillae which consist of an epithelial surface and a mesenchymal core.^^'^ '^ Shh plays very similar roles in tooth and papillae with respect to both early patterning and later epithelial-mesenchymal interactions. Murine in situ studies have shown Shh, along with Ptchl and GUI, are expressed diffusely in the developing tongue during early stages of embryogenesis (Fig. 10).^^^'^^^ Subsequently, Shk Bmp2 and Bmp4 are localized to discrete regions of the anterior tongue epithelium which correspond to where fungiform papillae will ultimately form (Fig. 10).^ ' Ptchl is expressed in the mesenchvme underlying 5;6/^-expressing epithelium and in areas surrounding the fungiform papillae. These data suggest that Shh specifies the locations of the fungiform papillae and is involved in epithelial-mesenchymal interactions and may play a role in papillary spacing. Shh may also play a role in directing the morphogenesis of the papillae by limiting the extent of papillary growth, both in the placodal epithelium and in the mesenchyme.^^ Inhibiting Shh signal transduction with cyclopamine or 5E1 results in significantly enlarged papillae and the development of papillae in ectopic regions, the interplacodal areas.^^'^^ This suggests that Shh normally signals to Ptchl in the mesenchvme of the periplacodal regions that result in repression of papillary development in these areas.

170


Figure 10. Gene expression in the developing tongue. BMP4 said Shh are both expressed in early developing tongues. A,B) Double X-g2l/in situ hybridization staining for BMP4 and Shh expression, respectively, in BMP4^^ tongues. BMP4^ staining is blue and Shh is purple. A) At E11.5, Shh is expressed broadly in the tongue in a different pattern than BMP4 , which is restricted to the midline mesenchyme. B) At El2.5, no difference in BMP4^^ and Shh expression is detected. C) Single image plane confocal micrograph of double-label immunohistochemistry for P-galactosidase (red) and Shh (green) in E13.5 BMP4 fungiform placodes, with a corresponding Nomarski image overlaid in (D). BMP4^^Sind Shh are present in all of the same cells in the developing papillary placodes. BMP4 is present in the cell nuclei because it carries a nuclear localization signal. Shh is present along the cell surface. Scale bars: A,B 500 |Xm; C,D 20 |Xm. Pictures courtesy Dr. Thomas Finger and Dr. Joshua Hall, and reproduced, with permission, from reference 89.

Submandibular Gland Development and Shh T h e submandibular glands (SMG) are located on each side of the lower jaw^ and produce saliva consisting of mucin, salts and amylase. Embryonic development of these glands depends on epithelial-mesenchymal interactions and branching. T h e stages of S M G development are classified as prebud, pseudoglandular, canalicular and terminal bud; each stage is delineated by particular morphological characteristics.'"^^ Initially, there is downv^rard grovs^th of the oral epithelium into the mesenchyme forming a bulb like structure.' '^ Branching morphogenesis subsequently occurs as a result of cellular interactions.''^' T h e Shh pathway is important in S M G development. T h e expression of Shh, Ptchl, Smo and Gli3 are localized to the epithelia at all stages of S M G development.''"^ T h e requirement for Shh in branching of S M G has been shown by knockout of the Shh gene and biochemical inhibition of Shh transduction. Evaluation of the S M G in E l 8 . 5 Shh" mice reveals tiny dysplastic structures with decreased branching. Decreased branching is also produced when cyclopamine is administered to cell culture explants of S M G primordia.''^ W h e n Fgf8 is added to the media of these explants the branching morphology is rescued suggesting that FgfS in the S M G is downstream of S h h . ' ' ^ It is also of interest that the expression o^ Ptchl, Smo and GUI in S M G are different from other organs in which branching occurs. Although Shh is expressed in the epithelium of S M G and other branched organs, Ptchl, Smo and Gli are expressed in the epithelium of S M G but expressed in the underlying mesenchyme of other branching organs."^ It remains unclear why the S M G is different in this respect.

Sonic Hedgehog Signaling in Craniofacial

Development

171

Future Directions One third of birth defects involve the craniofacial region. Fortunately, the past decade has witnessed exponential growth in our knowledge of craniofacial development due to the integration of data from a number of disciplines. A range of experimental strategies and abundant clinical data demonstrate the requirement for Shh signaling in normal craniofacial morphogenesis. We are just at the beginning of our journey into uncovering the many developmental complexities of the craniofacial region. Delineating the molecules and their actions involved in craniofacial development will continue to provide insight into the etiologies of congenital disorders that are now regarded as idiopathic. This knowledge will also provide novel treatment methodologies and more accurate diagnostic methods in both prenatal and postnatal periods.

Acknowlegefnents Victoria, Maria, and Jaron Cordero; Dilip, Deval and Anjali Tapadia; and I. Merkatz for their support.

References 1. Homer. (850 B.C.). T h e Odyssey. 2. Belloni E, Muenke M , Roessler E et al. Identification of Sonic hedgehog as a candidate gene responsible for holoprosencephaly. Nat Genet 1996; 14:353-6. 3. Schimmenti LA, de la Cruz J, Lewis RA et al. Novel mutation in sonic hedgehog in nonsyndromic colobomatous microphthalmia. A m J Med Genet 2 0 0 3 ; 116A:215-21. 3a. Jeff Dee. Unigames, 1999. http://www.io.com/unigames/jeffdee.html. 4. Cohen Jr M M . Problems in the definition of holoprosencephaly. Am J Med Genet 2 0 0 1 ; 103:183-7. 5. Demyer W , Zeman W , Palmer C G . T h e face predicts the brain: Diagnostic significance of median facial anomalies for holoprosencephaly (Arhinencephaly). Pediatrics 1964; 34:256-63. 6. O M I M . Online Mendelian Inheritance in Man, O M I M . Baltimore, Bethesda, M D : McKusickNathans Institute for Genetic Medicine, Johns Hopkins University and National Center for Biotechnology Information, National Library of Medicine, 2000. http://www.ncbi.nlm.nih.gov/omim/. 7. Nanni L, M i n g JE, Bocian M et al. T h e mutational spectrum of the sonic hedgehog gene in holoprosencephaly: S H H m u t a t i o n s cause a significant p r o p o r t i o n of autosomal d o m i n a n t holoprosencephaly. H u m Mol Genet 1999; 8:2479-88. 8. Roessler E, Belloni E, Gaudenz K et al. Mutations in the C-terminal domain of Sonic Hedgehog cause holoprosencephaly. H u m Mol Genet 1997; 6:1847-53. 9. Nanni L, M i n g JE, D u Y et al. S H H mutation is associated with solitary median maxillary central incisor: A study of 13 patients and review of the Hterature. A m J Med Genet 2 0 0 1 ; 102:1-10. 10. WalUs D , Muenke M . Mutations in holoprosencephaly. H u m Mutat 2000; 16:99-108. 11. Ming JE, Kaupas ME, Roessler E et al. Mutations in P A T C H E D - 1 , the receptor for S O N I C H E D G E H O G , are associated with holoprosencephaly. H u m Genet 2002; 110:297-301. 12. Marini M , Cusano R, De Biasio P et al. Previously undescribed nonsense mutation in S H H caused autosomal dominant holoprosencephaly with wide intrafamilial variability. A m J Med Genet 2003; 117A: 112-5. 13. Traiffort E, Dubourg C, Faure H et al. Functional characterization of sonic hedgehog mutations associated with holoprosencephaly. J Biol Chem 2004; 279:42889-97. 14. Briscoe J, Chen Y, Jessell T M et al. A hedgehog-insensitive form of patched provides evidence for direct long-range morphogen activity of sonic hedgehog in the neural tube. Mol Cell 2 0 0 1 ; 7:1279-91. 15. Roessler E, D u YZ, MuUor JL et al. Loss-of-function mutations in the h u m a n GLI2 gene are associated with pituitary anomalies and holoprosencephaly-like features. Proc Natl Acad Sci USA 2003; 100:13424-9. 16. Johnson RL, Milenkovic L, Scott M P . In vivo functions of the patched protein: Requirement of the C terminus for target gene inactivation but not Hedgehog sequestration. Mol Cell 2000; 6:467-78. 17. Boutet N , Bignon YJ, Drouin-Garraud V et al. Spectrum of P T C H l mutations in French patients with GorUn syndrome. J Invest Dermatol 2003; 121:478-81. 18. Lam C W , Leung CY, Lee KG et al. Novel mutations in the P A T C H E D gene in basal cell nevus syndrome. Mol Genet Metab 2002; 76:57-61. 19. Gorlin RJ. Gorlin (nevoid basal cell carcinoma) syndrome. In: Gorlin RJ, Cohen M M , Hennekam R C M , eds. Syndromes of the Head and Neck. Oxford: Oxford University Press, 2 0 0 1 .

172

Hedgehog-Gli

Signaling in Human

Disease

20. Lacombe D , Chateil JF, Fontan D et al. Medulloblastoma in the nevoid basal-cell carcinoma syndrome: Case reports and review of the literature. Genet Couns 1990; 1:273-7. 2 1 . Kalfif-Suske M , Wild A, T o p p J et al. Point mutations throughout the GLI3 gene cause Greig cephalopolysyndactyly syndrome. H u m Mol Genet 1999; 8:1769-77. 22. Kang S, Graham Jr J M , Olney A H et al. GLI3 frameshift mutations cause autosomal dominant Pallister-Hall syndrome. N a t Genet 1997a; 15:266-8. 2 3 . Shin SH, Kogerman P, Lindstrom E et al. GLI3 mutations in human disorders mimic Drosophila cubitus interruptus protein functions and localization. Proc Natl Acad Sci USA 1999; 96:2880-4. 24. Duncan PA, Klein R M , Wilmot PL et al. Greig cephalopolysyndactyly syndrome. Am J Dis Child 1979; 133:818-21. 2 5 . Greig D . Oxycephaly. Edinburgh Medical Journal 1926; 33:189-218. 26. Kang S, Rosenberg M , Ko V D et al. Gene structure and alleUc expression assay of the human GLI3 gene. H u m Genet 1997b; 101:154-7. 27. Vortkamp A, Gessler M , Grzeschik K H . GLI3 zinc-finger gene interrupted by translocations in Greig syndrome families. Nature 1991; 352:539-40. 28. Wild A, Kalff-Suske M , Vortkamp A et al. Point mutations in human GLI3 cause Greig syndrome. H u m Mol Genet 1997; 6:1979-84. 29. Kelley RL, Roessler E, Hennekam RC et al. Holoprosencephaly in RSH/Smith-Lemli-Opitz syndrome: Does abnormal cholesterol metabolism affect the function of Sonic Hedgehog? Am J Med Genet 1996; 66:478-84. 30. Muenke M, Beachy PA. Genetics of ventral forebrain development and holoprosencephaly. Curr O p i n Genet Dev 2000; 10(3):262-9. 3 1 . Worthington S, Goldblatt J. Smith-Lemli-Opitz syndrome: further delineation of the phenotype. Clin Dysmorphol 1997; 6(3):263-6. 32. H o n d a A, T i n t GS, Salen G et al. Defective conversion of 7-dehydrocholesterol to cholesterol in cultured skin fibroblasts from Smith-Lemli-Opitz syndrome homozygotes. J Lipid Res 1995; 36:1595-601. 33. Shefer S, Salen G, Batta AK et al. Markedly inhibited 7-dehydrocholesterol-dclta 7-reductase activity in liver microsomes from Smith-Lemh-Opitz homozygotes. J CHn Invest 1995; 96:1779-85. 34. Zeng X, Goetz JA, Suber LM et al. A freely difiiisiblc form of Sonic hedgehog mediates long-range signalling. Nature 2 0 0 1 ; 411:716-20. 35. Roux C, Wolf C, Mulliez N et al. Role of cholesterol in embryonic development. Am J Clin Nutr 2000; 71:1270S-9S. 36. Edison RJ, Muenke M . Central nervous system and limb anomalies in case reports of first-trimester statin exposure. N Engl J Med 2004; 350:1579-82. 37. C D C . Fetal Alcohol Information. Centers for Disease Control 2004, (http://www.cdc.gov/ncbddd/ fas/fasask.htm). 38. Jones KL, Smith D W , Ulleland C N et al. Pattern of malformation in offspring of chronic alcoholic mothers. Lancet 1973; 1:1267-71. 39. Sampson P D , Streissguth AP, Bookstein FL et al. Incidence of fetal alcohol syndrome and prevalence of alcohol-related neurodevelopmental disorder. Teratology 1997; 56:317-26. 40. Young DL, Schneider RA, H u D et al. Genetic and teratogenic approaches to craniofacial development. Crit Rev Oral Biol Med 2000; 11(3):304-17. 4 1 . Ahlgren SC, Thakur V, Bronner-Fraser M . Sonic hedgehog rescues cranial neural crest from cell death induced by ethanol exposure. Proc Natl Acad Sci USA 2002; 99:10476-81. 42. Zachman RD, Grummer MA. T h e interaction of ethanol and vitamin A as a potential mechanism for the pathogenesis of Fetal Alcohol syndrome. Alcohol Clin Exp Res 1998; 22:1544-56. 4 3 . Stern RS, Rosa F, Baum C. Isotretinoin and pregnancy. J Am Acad Dermatol 1984; 10:851-4. 44. Monga M. Vitamin A and its congeners. Semin Perinatol 1997; 21:135-42. 4 5 . N a u H . Teratogenicity of isotretinoin revisited: Species variation and the role of all-trans-retinoic acid. J Am Acad Dermatol 2001; 45:S 183-7. AG. Helms JA, Kim C H , H u D et al. Sonic hedgehog participates in craniofacial morphogenesis and is down-regulated by teratogenic doses of retinoic acid. Dev Biol 1997; 187:25-35. 47. Hayhurst M , McConnell SK. Mouse models of holoprosencephaly. Curr O p i n Neurol 2003; 16:135-41. 48. Chiang C, Litingtung Y, Lee E et al. Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 1996; 383:407-13. 49. Goodrich LV, Milenkovic L, Higgins KM et al. Altered neural cell fates and medulloblastoma in mouse patched mutants. Science 1997; 277:1109-13. 50. Jeong J, Mao J, Tenzen T et al. Hedgehog signaling in the neural crest cells regulates the patterning and growth of facial primordia. Genes Dev 2004; 18:937-51.


173

5 1 . Milenkovic L, Goodrich LV, Higgins KM et al. Mouse patched 1 controls body size determination and limb patterning. Development 1999; 126:4431-40. 52. Cordero D , Milenkovic L, Tapadia M et al. Striking a balance between Sonic Hedgehog and Patched during craniofacial development. Abstract Annual Meeting of the American College of Medical Genetics. Dallas, Texas, 2005. 53. Goodrich LV, Jung D , Higgins KM et al. Overexpression of p t c l inhibits induction of Shh target genes and prevents normal patterning in the neural tube. Dev Biol 1999; 211:323-34. 54. H u i C C , Joyner AL. A mouse model of greig cephalopolysyndactyly syndrome: T h e extra-toesj mutation contains an intragenic deletion of the GU3 gene. N a t Genet 1993; 3:241-6. 55. Aoto K, Nishimura T , Eto K et al. Mouse GLI3 regulates Fgf8 expression and apoptosis in the developing neural tube, face, and limb bud. Dev Biol 2002; 251:320-32. 56. Ding Q, Motoyama J, Gasca S et al. Diminished Sonic hedgehog signaling and lack of floor plate differentiation in Gli2 mutant mice. Development 1998; 125:2533-43. 57. Lewis KE, Drossopoulou G, Paton IR et al. Expression of ptc and gli genes in talpid3 suggests bifurcation in Shh pathway. Development 1999; 126:2397-407. 58. Buxton P, Francis-West P H , Davey M G et al. Craniofacial development in the talpid3 chicken mutant. Differentiation 2004; 72:348-62. 59. Molotkov A, Duester G. Retinol/ethanol drug interaction during acute alcohol intoxication in mice involves inhibition of retinol metabolism to retinoic acid by alcohol dehydrogenase. J Biol C h e m 2002; 277:22553-7. 60. Schneider RA, H u D , Rubenstein JL et al. Local retinoid signaling coordinates forebrain and facial morphogenesis by maintaining FGF8 and S H H . Development 2 0 0 1 ; 128:2755-67. 6 1 . Hoffman L, Miles J, Avaron F et al. Exogenous retinoic acid induces a stage-specific, transient and progressive extension of Sonic hedgehog expression across the pectoral fin bud of zebrafish. Int J Dev Biol 2002; 46:949-56. 62. Keeler RF. Teratogenic compounds of Veratrum californicum (Durand) X. Cyclopia in rabbits produced by cyclopamine. Teratology 1970; 3:175-80. 63. Keeler RF. Livestock models of human birth defects, reviewed in relation to poisonous plants. J Anim Sci 1988; 66:2414-27. 64. Incardona JP, Gaffleld W , Kapur RP et al. T h e teratogenic Veratrum alkaloid cyclopamine inhibits sonic hedgehog signal transduction. Development 1998; 125:3553-62. 65. James LF. Teratological research at the USDA-ARS poisonous plant research laboratory. J N a t Toxins 1999; 8:63-80. GG. Binns W , Thacker EJ, James LF et al. A congenital cyclopiantype malformation in lambs. J Am Vet Med Assoc 1959; 134:180-3. G7. Binns W, James LF, Shupe JL et al. A congenital cyclopian-type malformation in lambs induced by maternal ingestion of a range plant, veratrum californicum. A m J Vet Res 1963; 24:1164-75. 68. Keeler RF, Binns W . Teratogenic compounds of Veratrum californicum (Durand). V. Comparison of cyclopian effects of steroidal alkaloids from the plant and structurally related compounds from other sources. Teratology 1968; 1:5-10. 69. Keeler RF. Teratogenic compounds of Veratrum californicum (Durand). VII. T h e structure of the glycosidic alkaloid cycloposine. Steroids 1969; 13:579-88. 70. Keeler RF. Cyclopamine and related steroidal alkaloid teratogens: Their occurrence, structural relationship, and biologic effects. Lipids 1978; 13:708-15. 7 1 . Gaffield W , Keeler RF. Induction of terata in hamsters by solanidane alkaloids derived from Solanum tuberosum. C h e m Res Toxicol 1996; 9:426-33. 72. Coventry S, Kapur RP, Siebert JR. Cyclopamine-induced holoprosencephaly and associated craniofacial malformations in the golden hamster: Anatomic and molecular events. Pediatr Dev Pathol 1998; 1:29-41. 7 3 . Cordero D , Marcucio R, H u D et al. Temporal perturbations in sonic hedgehog signaling elicit the spectrum of holoprosencephaly phenotypes. J CUn Invest 2004; 114:485-94. 74. Chen JK, Taipale J, Cooper M K et al. Inhibition of Hedgehog signaling by direct binding of cyclopamine to Smoothened. Genes Dev 2002; 16:2743-8. 75. Cooper MK, Porter JA, Young KE et al. Teratogen-mediated inhibition of target tissue response to Shh signaling. Science 1998; 280:1603-7. 7G. Lewis P M , D u n n M P , McMahon JA et al. Cholesterol modification of sonic hedgehog is required for long-range signaling activity and effective m o d u l a t i o n of signaling by P t c l . Cell 2 0 0 1 ; 105:599-612. 77. Kolf-Clauw M , Chevy F, Wolf C et al. Inhibition of 7-dehydrocholesterol reductase by the teratogen AY9944: A rat model for Smith-Lemli-Opitz syndrome. Teratology 1996; 54:115-25.

174

Hedgehog-Gli

Signaling in Human

Disease

78. Kolf-Clauw M , Chevy F, Siliart B et al. Cholesterol biosynthesis inhibited by BM 15.766 induces holoprosencephaly in the rat. Teratology 1997; 56:188-200. 79. Cohen Jr M M . T h e hedgehog signaling network. Am J Med Genet 2003; 123A:5-28. 80. Willnow T E , Hilpert J, Armstrong SA et al. Defective forebrain development in mice lacking gp330/megalin. Proc Natl Acad Sci USA 1996; 93:8460-4. 8 1 . Barth JL, Argraves W S . Cubilin and megalin: Partners in lipoprotein and vitamin metabolism. Trends Cardiovasc Med 2 0 0 1 ; 11:26-31. 82. Drake CJ, Fleming PA, Larue AC et al. Differential distribution of cubilin and megalin expression in the mouse embryo. Anat Rec 2004; 277A: 163-70. 83. McCarthy RA, Barth JL, Chintalapudi M R et al. Megalin functions as an endocytic sonic hedgehog receptor. J Biol Chem 2002; 277:25660-7. 84. Christensen EI, Moskaug J O , V o r u m H et al. Evidence for an essential role of megalin in transepithelial transport of retinol. J A m Soc Nephrol 1999; 10:685-95. 85. McCarthy RA, Argraves W S . Megalin and the neurodevelopmental biology of sonic hedgehog and retinol. J Cell Sci 2003; 116:955-60. 86. Birn H , Vorum H , Verroust PJ et al. Receptor-associated protein is important for normal processing of megalin in kidney proximal tubules. J Am Soc Nephrol 2000; 11:191-202. 87. Ericson J, Morton S, Kawakami A et al. Two critical periods of Sonic Hedgehog signaling required for the specification of motor neuron identity. Cell 1996; 87:661-73. 88. Cobourne M T , Hardcastle Z, Sharpe PT. Sonic hedgehog regulates epithelial proliferation and cell survival in the developing tooth germ. J Dent Res 2 0 0 1 ; 80:1974-9. 89. Hall JM, Bell ML, Finger T E . Disruption of sonic hedgehog signaling alters growth and patterning of lingual taste papillae. Dev Biol 2003; 255:263-77. 90. Dillon C P , Sandy P, Nencioni A et al. RNAi as an experimental and therapeutic tool to study and regulate physiological and disease processes. Annu Rev Physiol 2004. 9 1 . Mello C C , Conte Jr D . Revealing the world of RNA interference. Nature 2004; 431:338-42. 92. Mocellin S, Provenzano M . RNA interference: Learning gene knock-down from cell physiology. J Transl Med 2004; 2:39. 9 3 . Wadhwa R, Kaul SC, Miyagishi M et al. Vectors for RNA interference. Curr Opin Mol Ther 2004; 6:367-72. 94. Helms JA, Schneider RA. Cranial skeletal biology. Nature 2003; 423:326-31. 95. Le Douarin N M , Creuzet S, Couly G et al. Neural crest cell plasticity and its limits. Development 2004; 131:4637-50. 96. S p e m a n n H a M , H . t b e r I n d u k t i o n von E m b r y o n a n l a g e n durch I m p l a n t a t i o n artfremder Organisatoren. Roux' Arch f Entw mech 1924; 100:599-638. 97. Tabata T, Takei Y. Morphogens, their identification and regulation. Development 2004; 131:703-12. 98. Chen Y, Struhl G. Dual roles for patched in sequestering and transducing Hedgehog. Cell 1996; 87:553-63. 99. Lee CS, Buttitta L, Fan C M . Evidence that the WNT-inducible growth arrest-specific gene 1 encodes an antagonist of sonic hedgehog signaling in the somite. Proc Natl Acad Sci USA 2001; 98:11347-52. 100. Ruiz i Altaba A, Nguyen V, Palma V. T h e emergent design of the neural tube: Prepattern, S H H morphogen and GLI code. Curr O p i n Genet Dev 2003; 13:513-21. 101. Goodrich LV, Scott M P . Hedgehog and patched in neural development and disease. Neuron 1998; 21:1243-57. 102. Briscoe J, Pierani A, Jessell T M et al. A homeodomain protein code specifies progenitor cell identity and neuronal fate in the ventral neural tube. Cell 2000; 101:435-45. 103. Timmer JR, W a n g C, Niswander L. BMP signaling patterns the dorsal and intermediate neural tube via regulation of homeobox and helix-loop-helix transcription factors. Development 2002; 129:2459-72. 104. Crossley P H , Martinez S, Ohkubo Y et al. Coordinate expression of Fgf«8, Otx2, Bmp4, and Shh in the rostral prosencephalon during development of the telencephalic and optic vesicles. Neuroscience 2 0 0 1 ; 108:183-206. 105. M u e n k e M , C o h e n Jr M M . G e n e t i c a p p r o a c h e s to u n d e r s t a n d i n g brain d e v e l o p m e n t : Holoprosencephaly as a model. Ment Retard Dev Disabil Res Rev 2000; 6:15-21. 106. Marcucio R, Cordero D , H u D et al. Molecular interactions coordinating development of the forebrain and face. Dev Biol 2005; in press. 107. Knoetgen H , Teichmann U, Kessel M . Head-organizing activities of endodermal tissues in vertebrates. Cell Mol Biol (Noisy-le-grand) 1999; 45:481-92. 108. H u D , Marcucio RS, Helms JA. A zone of frontonasal ectoderm regulates patterning and growth in the face. Development 2003; 130:1749-58.

Sonic Hedgehog Signaling in Craniofacial

Development

175

109. Mina M , Kollar EJ. T h e induction of odontogenesis in nondental mesenchyme combined with early murine mandibular arch epithelium. Arch Oral Biol 1987; 32:123-7. 110. Lumsden A G . Spatial organization of the epithelium and the role of neural crest cells in the initiation of the mammalian tooth germ. Development 1988; 103(Suppl):155-69. 111. Francis-West P, Ladher R, Barlow A et al. Signalling interactions during facial development. Mech Dev 1998; 75:3-28. 112. Rice R, Spencer-Dene B, Connor EC et al. Disruption of Fgfl0/Fgfr2b-coordinated epithelial-mesenchymal interactions causes cleft palate. J Clin Invest 2004; 113:1692-700. 113. Cobourne M T , Miletich I, Sharpe PT. Restriction of sonic hedgehog signalling during early tooth development. Development 2004; 131:2875-85. 114. Hall JM, Hooper JE, Finger TE. Expression of sonic hedgehog, patched, and Glil in developing taste papillae of the mouse. J Comp Neurol 1999; 406:143-55. 115. JaskoU T , Melnick M . Submandibular gland morphogenesis: Stê-specific expression of TGF-alpha/ EOF, IGF, TGF-beta, T N F , and IL-6 signal transduction in normal embryonic mice and the phenotypic effects of TGF-beta2, TGF-beta3, and EGF-r null mutations. Anat Rec 1999; 256:252-68. 116. Kashimata M , Sayeed S, Ka A et al. T h e ERK-1/2 signaling pathway is involved in the stimulation of branching morphogenesis of fetal mouse submandibular glands by E G F . Dev Biol 2 0 0 0 ; 220:183-96. 117. Jaskoll T, Leo T, Witcher D et al. Sonic hedgehog signaling plays an essential role during embryonic salivary gland epithelial branching morphogenesis. Dev Dyn 2004; 229:722-32. 117a.Cordero D , Schneider RA, Helms JA. Morphogenesis of the Face. In: Lin O , Jane, eds. In Craniofacial Surgery: Science and Surgical Technique. Philadelphia: W B Saunders Company 2001:75-83. 118. Lacbawan FL, Muenke M. Central nervous system embryogenesis and its failures. Pediatr Dev Pathol 2002; 5:425-47. 119. Monuki ES, Walsh CA. Mechanisms of cerebral cortical patterning in mice and humans. Nat Neurosci 2 0 0 1 ; 4(Suppl): 1199-206. 120. Roessler E, Muenke M . MidUne and laterality defects: Left and right meet in the middle. Bioessays 2 0 0 1 ; 23:888-900. 121. Kessaris N , Jamen F, Rubin LL et al. Cooperation between sonic hedgehog and fibroblast growth factor/MAPK signalUng pathways in neocortical precursors. Development 2004; 131:1289-98. 122. Marklund M , Sjodal M , Beehler BC et al. Retinoic acid signalling specifies intermediate character in the developing telencephalon. Development 2004; 131:4323-32. 123. Palma V, Ruiz i Altaba A. Hedgehog-GLI signaling regulates the behavior of cells with stem cell properties in the developing neocortex. Development 2004; 131:337-45. 124. Le Douarin N M , Dupin E. Multipotentiality of the neural crest. Curr O p i n Genet Dev 2003; 13:529-36. 125. Noden D M . Origins and patterning of craniofacial mesenchymal tissues. J Craniofac Genet Dev Biol Suppl 1986; 2:15-31. 126. Trainor PA, Melton KR, Manzanares M . Origins and plasticity of neural crest cells and their roles in jaw and craniofacial evolution. Int J Dev Biol 2003; 47:541-53. 127. Kulesa P, EUies DL, Trainor PA. Comparative analysis of neural crest cell death, migration, and function during vertebrate embryogenesis. Dev Dyn 2004; 229:14-29. 128. T r u m p p A, Depew MJ, Rubenstein JL et al. Cremediated gene inactivation demonstrates that FGF8 is required for cell survival and patterning of the first branchial arch. Genes Dev 1999; 13:3136-48. 129. Schneider RA, Helms JA. T h e cellular and molecular origins of beak morphology. Science 2003; 299:565-8. 130. Tucker AS, Lumsden A. Neural crest cells provide species-specific patterning information in the developing branchial skeleton. Evol Dev 2004; 6:32-40. 131. H u D , Helms JA. T h e role of sonic hedgehog in normal and abnormal craniofacial morphogenesis. Development 1999; 126:4873-84. 132. Abzhanov A, Helms JA et al. Cross-regulatory interactions between Fgf8 and Shh in chick frontonasal primordium. In revision. 133. M o R, Freer A M , Zinyk D L et al. Specific and redundant functions of Gli2 and Gli3 zinc finger genes in skeletal patterning and development. Development 1997; 124:113-23. 134. Ericson J, M u h r J, Placzek M et al. Sonic hedgehog induces the differentiation of ventral forebrain neurons: a c o m m o n signal for ventral patterning within the neural tube. Cell 1995; 81(5):747-56. 135. Ferguson M W . Palate development. Development 1988; 103(Suppl):41-60. 136. De Moerlooze L, Spencer-Dene B, Revest J et al. An important role for the Illb isoform of fibroblast growth factor receptor 2 (FGFR2) in mesenchymal-epithelial signalling during mouse organogenesis. Development 2000; 127:483-92.

176


137. Igarashi M, Finch PW, Aaronson SA. Characterization of recombinant human fibroblast growth factor (FGF)-IO reveals functional similarities with keratinocyte growth factor (FGF-7). J Biol Chem 1998; 273:13230-5. 138. Vieira AR, Karras JC, Orioli IM et al. Genetic origins in a South American clefting population. Clin Genet 2002; 62:458-63. 139. Peters H, Balling R. Teeth. Where and how to make them. Trends Genet 1999; 15:59-65. 140. Tucker AS, Sharpe PT. Molecular genetics of tooth morphogenesis and patterning: The right shape in the right place. J Dent Res 1999; 78:826-34. 141. Jernvall J, Thesleff I. Reiterative signaling and patterning during mammalian tooth morphogenesis. Mech Dev 2000; 92:19-29. 142. Farbman AI, Mbiene JP. Early development and innervation of taste bud-bearing papillae on the rat tongue. J Comp Neurol 1991; 304:172-86. 143. Jung HS, Oropeza V, Thesleff I. Shh, Bmp-2, Bmp-4 and Fgf-8 are associated with initiation and patterning of mouse tongue papillae. Mech Dev 1999; 81:179-82. 144. Mistretta CM, Liu HX, Gaffield W et al. Cyclopamine and jervine in embryonic rat tongue cultures demonstrate a role for Shh signaling in taste papilla development and patterning: Fungiform papillae double in number and form in novel locations in dorsal lingual epithelium. Dev Biol 2003; 254:1-18. 145. Melnick M, Chen H, Zhou Y et al. Embryonic mouse submandibular salivary gland morphogenesis and the TNF/TNF-Rl signal transduction pathway. Anat Rec 2001; 262:318-30. 146. Cutler LS, Gremski W. Epithelial-mesenchymal interactions in the development of salivary glands. Grit Rev Oral Biol Med 1991; 2:1-12. 147. Helms JA, Cordero D, Tapadia MD. New insights into craniofacial morphogenesis. Development 2005; 132:851-61.

CHAPTER 14

Important Role of Shh Controlling Gli3 Functions during the Dorsal-Ventral Patterning of the Telencephalon Jun Motoyama* and Kazushi Aoto Abstract

T

he dorsal-ventral patterning of the telencephalon is crucial for normal brain function because it determines the proportion of two different basic types of neurons: glutamatergic excitatory neurons and GABAergic inhibitory neurons. The secreted protein sonic hedgehog (Shh) is required for ventral cell specification, whereas the zinc finger transcription factor Gli3 seems to be important for dorsal cell type specification. Recent studies suggest how both Gli3 and Shh control the normal proportion of dorsal and ventral cell types to generate appropriate tissue size and shape. These observations may offer new insights into our understanding of the graded function of Shh during brain development.

Introduction Embryogenesis involves a complex coordination of cell fate specification, proliferation and differentiation that is determined by the actions of a legion of genes. The sonic hedgehog (Shh) signaling cascade is important in many developmental processes of the central nervous system, and dysfunctions in this pathway can lead to birth defects and brain tumors.^ Although the roles of the individual genes of Shh signaling cascade in each of these developmental processes have started to be defined, how these gene functions combine to translate the dynamic assembly of cells into tissues remains unresolved. Recent genetic studies using combined mutant mice have provided new insights into how several genes control the normal proportion of different cell types to generate appropriate tissue size and shape. Here, we review the role of Shh and Gli3 during dorsal-ventral patterning of the mouse telencephalon revealed by the analyses using not only Gli3 and Shh single mutants but also Gli3/Shh double mutant mice.

Shh Signaling and Dorsal-Ventral Patterning of the Telencephalon Sonic hedgehog (Shh) is a secreted protein that controls the patterning and proliferation of many tissues in developing human and mouse embryos. Many of the components that mediate the Shh signaling in these embryos were first discovered in their homologues affecting early embryonic development in Drosophila. In Drosophila, hedgehog (Hh) is a secreted protein that is required for pattern formation during many processes of fly development. Two transmembrane proteins, patched (Ptc) and smoothened (Smo), are involved in the reception of the Hh signals '^ and a transcription factor, cubitus interrupts (Ci), is the ultimate transducer of Hh *Corresponding Author: Jun Motoyama—Molecular Neuropathology Group, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan. Email: [email protected]


178


repressor Ci

Proteolytic cleavage \

N|||||||||||||[^

activator^ ^ ^ ^ ^ P C

1377 AA

Proteolytic cleavage

GH2

HHIBIIIUFDJ

m

• • '^''AA

Proteolytic cleavage

Figure 1. Schematic diagram of Ci/Gli family protein structure. All members have the zincfingerdomain (ZFD) and the activator domain in the C-termini (gray). Ci/Gli proteins, except for Gli 1, have the repressor domain in the N-termini (black) and act as repressors after processing in the proteolytic cleavage site (arrow). signaling inside target cells. ^^ In the absence of Hh, Ptc inhibits Smo function, allowing proteolytic cleavage of Ci which then suppresses the transcription of target genes. ^^ However, when Hh binds to Ptc, the repressive action of Ptc on Smo is blocked, Ci proteolysis is inhibited, and Ci is converted into a transcriptional activator. Several features of Hh signal transduction are evolutionarily conserved in vertebrates, and when this pathway is impaired birth defects and tumorigenesis result. Shh is encoded by one of three mammalian orthologous genes to Hhy desert hedgehog, Indian hedgehog 2nd sonic hedgehog. Mice and humans have at least three Ci homologues (G7/7, Gli2y and Gli3) which encode transcription factors mediating Shh signaling. ' Of these, Gli2 and Gli3 can be proteolytically processed and seem to have transcriptional activator and suppressor functions ^'^^ (Fig. 1). Mouse Gli genes were introduced in theflywing disc to examine whether the processing of mouse Glis can be regulated by Drosophila Hh signaling. Glil and Gli2 can function as activators of Hh signaling in theflydisc. Gli2 and Gli3 are processed in theflywing disc, but the processing of Gli2 is observed even within the posterior compartment expressing Hh signaling, suggesting that it is not regulated by Hh signaling. On the other hand, the proteolytic cleavage of Gli3 is regulated by Hh signaling in a similar manner to how Ci is regulated by Hh in Drosophila} Glil seems to have only an activator function^^ (Fig. 1). Also in the developing limb buds in mouse and chick embryos, proteolytic cleavage of Gli3 is observed and its distribution appears to be regulated by Shh from the posterior region of the limb bud named the ZPA (zone of polarizing activity). ^^ Thus, not only in Drosophila but also in mouse and chick, proteolytically processed Gli3 appears to be a specific antagonist of Shh signaling. The dorsal-ventral patterning of the telencephalon determines the proportion of two different types of basic neurons, glutamatergic excitatory neurons and GABAergic inhibitory neurons, which are necessary for normal telencephalon function. Fibroblast growth factor 8 (Fgf8) is an essential regulator for the dorsal-ventral patterning of the telencephalon. At the head fold stage, around 8.0 dpc, the presumptive forebrain is determined and the cells in the anterior-most aspect start expressing FgfB. Expressing cells in the anterior neural ridge are important for outgrowth of the telencephalon territory in the forebrain region (Fig. 2A,B). Depending on the

Important Role ofShh Controlling Gli3 Functions

179

Figure 2. Distribution of FgfB expression during telencephalon development. Scanning electron micrographs of embryonic telencephalon isolated from different stages during embryogenesis. Green area indicates the presumptive telencephalon in frontal (A, C, E, H) and lateral views (B, D, F, I). Purple area indicates the domain ofFgfS expression. G, J) Transverse section at yellow line in (F) and (I) are shown. Mesenchymal cells and surface ectoderm are removed in (E, F, H, I). E: eye; MGE: medial ganglionic eminence; LGE: lateral ganglionic eminence; CTX: cortex. Scale bars: 0.1 mm in (A, B, C, D); 0.2 mm in (E, F, H); 0.5 mm in (I). effect of the Fgf8, the surrounding neural plate grows rapidly and forms a pair of brain vesicles that becomes the presumptive telencephalon^^'"^^ (Fig. 2 C - G ) . T h e dorsal part of the telencephalon develops into the neocortex and the hippocampus, mainly generating excitatory glutamatergic projection neurons. T h e ventral telencephalon develops into lateral and medial ganglionic eminences (Fig. 2H-J). T h e majority of the neurons generated in the ganglionic

180


Figure 3. Distribution o^Shh, Gli3 and FgfB mRNA in the telencephalon of wild type mice at 11.5 dpc (A, B, C) and schematic diagram of genes expression in the telencephalon (D). Frontal views of the telencephalon after whole mount in situ hybridization are shown in (A-C). Shh is expressed in the ventral midline, while Gli3 expression is in the presumptive cortex (arrow heads in A, B). Strong FgfB expression is detected in the intermediate zone of the telencephalon (arrow heads in C). In the schematic diagram each color code indicates the expression of regional brain marker; Shh (green), Gli3 (pink), Nkx2.1 (red), Dlx2 (yellow) and Emxl/2 (blue). MGE: medial ganglionic eminence; LGE: lateral ganglionic eminence; CTX: cortex. eminences are inhibitory GABAergic interneurons, which tangentially migrate to the neocortex and make a circuit with excitatory projection neurons. Thus, this dorsal-ventral patterning is important to maintain the proper balance between two different types of basic excitatory and inhibitory neurons. Shh and Gli3 are required for this dorsal-ventral patterning of the telencephalon. Shh is expressed in the ventral midline and Gli3 is in the dorsal telencephalon*^^ (Fig. 3A,B,D). Shh is thought to be essential for directing the formation of the ventral telencephalon, because the medial and lateral ganglionic eminences are severely affected and a single forebrain vesicle develops with missing midline structures in Shh-/- mouse embryos^^ (Fig. 4B). However, the severe brain malformations observed in Shh-/- mutants are restored in Gli3/Shh double mutant embryos^^'^ '^^ (Fig. 4C). This finding suggests that the phenotype in Shh-/- single mutant mice includes secondary defects. Moreover, the restored phenotype also suggests that Gli3 function is involved in the secondary defects observed in Shh-/- mutant mice (Fig. 4 B ) . The important question is how Gli3 functions with Shh to regulate the normal proportion of different cell types in order to generate appropriate tissue size and shape during telencephalon development.

Important Role ofShh Controlling Gli3 Functions

WT

>V

B Shh-I-

C

0113-1-

181

D Gli3-I-; Shh-I-

f^ s 3 £ iS Dorsal

Li • ^

^

•C CM

Ventral

^

^

^5 5 • Paxf ^Bmps/Wnts—• EmscfA Shh
/W«/^ expression by Gli3 appears to be essential in determining dorsal cell types. Importantly, in embryonic neuronal progenitors isolated from the dorsal telencephalon, the inhibition of Bmp signaling was reported to be sufficient to determine ventral cell types shown by using cell culture system. ^'^ Blocking Bmp signaling in the neuronal progenitors isolated from embryonic dorsal telencephalon with a dominant-negative Bmp receptor lb can induce difFerentiation into inhibitory interneurons, even if Shh signaling is also blocked by cyclopamine, an inhibitor of Smoothened function.^^ Thus, for dorsal progenitor cells, inhibition of Bmp signaling is necessary and sufficient for the induction of a dorsal-to-ventral shift in progenitor cell identity. Based on the mutant phenotype analyses and ectopic expression experiments, we already know that Gli3 is required for the activation o£Bmp genes expression^'^ and that the processed form of Gli3 can induce a ventral-to-dorsal shift in progenitor cell identity.^ '^^ These findings suggest that Shh is required for inhibition of dorsal cell specification by inhibiting Bmp signaling via the conversion of Gli3 repressors into activators only in the ventral region. The spatial and temporal distribution of Gli3 repressor form regulated by the activity gradient of Shh may therefore control the normal proportion between dorsal and ventral territories. Because most of the results summarized here are based on analyses of tissues from mutant animals, many unanswered questions about the functions of Gli3 and Shh proteins remain. Confirmation of the cellular functions of Gli3 is essential to extend these results to the molecular level. The results presented here indicate that the proper balance between Shh and Gli3 functions is important for normal dorsal-ventral patterning of the neural tube and brain. As Shh and Gli3 inhibit each other, when one is missing severe malformation results as the remaining one expands to the other's territory. Recent findings using double mutant studies revealed that both Shh and Gli3 regulate only a set of specific cell types during neural tube and brain development. Moreover, when both are absent many cell types located in the intermediate zone of the brain and neural tube develop normally, indicating that Shh and Gli3 can only modulate the dorsal-ventral patterning established by other signaling pathways. Bmp/Wnt signaling and their antagonists are essential for neural plate development and its specification before the expression of Shh in the neural plate cells has been established.^^'^^ It is likely that other signaling pathways participate in dorsal-ventral patterning in the telencephalon, but their precise roles and whether they act in concert or in parallel with Shh signaling requires further investigation.

Acknowledgements We thank Drs. Adrian Moore and Bonnie Lee Madeleine for helpful discussion and critical reading of the manuscript.

References 1. Riddle RD, Johnson RL, Laufer E et al. Sonic hedgehog mediates the polarizing activity of the ZPA. Cell 1993; 75:1401-1416. 2. Krauss S, Concordet JP, Ingham PW. A functionally conserved homolog of the Drosophila segment polarity gene hh is expressed in tissues with polarizing activity in zebrafish embryos. Cell 1993; 75:1431-1444. 3. Echelard Y, Epstein DJ, St-Jacques B et al. Sonic hedgehog, a member of a family of putative signaling molecules, is implicated in the regulation of CNS polarity. Cell 1993; 75:1417-1430. 4. Roelink H, Porter JA, Chiang C et al. Floor plate and motor neuron induction by vhh-1, a vertebrate homolog of Hedgehog expressed by the notochord. Cell 1994; 76:761-775. 5. Porter JA, Young KE, Beachy PA. Cholesterol modification of hedgehog signaling proteins in animal development. Science 1996; 274:255-259.

Important

Role ofShh

Controlling

Gli3 Functions

185

6. O r o AE, Higgins K M , H u Z et al. Basal cell carcinomas in mice overexpressing sonic hedgehog. Science 1997; 276:817-821. 7. Fan H, O r o AE, Scott M P et al. Induction of basal cell carcinoma features in transgenic human skin expressing Sonic Hedgehog. Nature Med 1997; 3:788-792. 8. Chen Y, Struhl G. Dual roles for patched in sequestering and transducing hedgehog. Cell 1996; 87:553-563. 9. Quirk J, van den Heuvel M, Henrique D et al. T h e smoothened gene and hedgehog signal transduction in Drosophila and vertebrate development. Cold Spring Harb Symp Q u a n t Biol 1997; 62:217-226. 10. Methot N , Basler K. An absolute requirement for Cubitus interruptus in Hedgehog signaling. Development 2 0 0 1 ; 128:733-742. 11. Aza-Blanc P, Ramirez-Weber F, Laget M et al. Proteolysis that is inhibited by hedgehog targets cubitus interruptus protein to the nucleus and converts it to a repressor. Cell 1997; 89:1043-1053. 12. Lee J, Piatt KA, Censullo P et al. Glil is a target of Sonic hedgehog that induces ventral neural tube development. Development 1997; 124:2537-2552. 13. Hynes M, Stone D M , Dowd M et al. Control of cell pattern in the neural tube by the zinc finger transcription factor and oncogene GUI. Neuron 1997; 19:15-26. 14. Park HL, Bai C, Piatt KA et al. Mouse Glil mutants are viable but have defects in S H H signaling in combination with a GH2 mutation. Development 2000; 127:1593-1605. 15. Hashimoto-Torii K, Motoyama J, H u i C C et al. Differential activities of Sonic hedgehog mediated by Gli transcription factors define distinct neuronal subtypes in the dorsal thalamus. Mech Dev 2003; 120(10):1097-1111. 16. Ding Q , Motoyama J, Gasca S et al. Diminished Sonic hedgehog signaling and lack of floor plate differentiation in Gli2 mutant mice. Development 1998; 125:2533-2543. 17. Matise M P , Epstein DJ, Park H L et al. Gli2 is required for induction of floor plate and adjacent cells, but not most ventral neurons in the mouse central nervous system. Development 1998; 125:2759-2770. 18. Aza-Blanc P, Lin H , Ruiz i Altaba A et al. Expression of the vertebrate Gli proteins in Drosophila reveals a distribution of activator and repressor activities. Development 2000; 127:4293-4301. 19. W a n g B, Fallon JF, Beachy PA. Hedghog-regulated processing of Gli3 produces an anterior/posterior repressor gradient in the developing vertebrate Umb. Cell 2000; 100(4):423-434. 20. Shimamura K, Rubenstein JL. Inductive interactions direct early regionalization of the mouse forebrain. Development 1997; 124(14):2709-2718. 2 1 . Storm EE, Rubenstein JL, Martin GR. Dosage of Fgf8 determines whether cell survival is positively or negatively regulated in the developing forebrain. Proc N a t l Acad Sci USA 2 0 0 3 ; 100(4):1757-1762. 22. Aoto K, Nishimura T , Eto K et al. Mouse GLI3 regulates FGF8 expression and apoptosis in the developing neural tube, face and limb bud. Dev Biol 2002; 251:320-332. 2 3 . Chiang C, Litingtung Y, Lee E et al. Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 1996; 383:407-413. 24. Litingtung Y, Chiang C. Specification of ventral neuron types is mediated by an antagonistic interaction between Shh and GU3. Nature Neurosci 2000; 3:979-985. 25. Rallu M, Machold R, Gariano N et al. Dorsal-ventral patterning is established in the telencephalon of mutants lacking both Gli3 and hedgehog signaling. Development 2002; 129:4963-4774. 26. Theil T , Alvarez-Bolado G, Walter A et al. Gli3 is required for Emx gene expression during dorsal telencephalon development. Development 1999; 126:3561-3571. 27. Tole S, Ragsdale C W , Grove EA. Dorsal-ventral patterning of the telencephalon is disrupted in the mouse mutant extra-toes Q). Dev Biol 2000; 217:254-265. 28. Johnson DR. Extra-toes: A new mutant gene causing multiple abnormalities in the mouse. J Embryol Exp M o r p h 1967; 17:543-581. 29. Kuschel S, Ruther U, Theil T . A disrupted balance between B m p / W n t and Fgf signaling underUes the ventralization of the Gli3 mutant telencephalon. Dev Biol 2003; 260(2):484-495. 30. Crossley P H , Martinez S, O h k u b o Y et al. Coordinate expression of F G F 8 , O t x 2 , Bmp4, and Shh in the anterior prosencephalon during development of the telencephalic and optic vesicles. N e u r o science 2 0 0 1 ; 108:183-206. 3 1 . Furuta Y, Piston D W , Hogan BL. Bone morphogenetic proteins (BMPs) as regulators of dorsal forebrain development. Development 1994; 124:2203-2212. 32. Theil T , Aydin S, Kosh S et al. W n t and Bmp signaling cooperatively regulate graded Emx2 expression in the dorsal telencephalon. Development 2002; 129(13):3045-3054. 33. Panchision D M , Pickel J M , Studer L et al. Sequential action of BMP receptor control neural precursor cell production and fate. Genes Dev 2 0 0 1 ; 15:2094-2110.

186


34. Persson M, Stamataki D, te Welscher P et al. Dorsal-ventral patterning of the spinal cord requires Gli3 transcriptional repressor activity. Genes Dev 2002; 16(22):2865-2878. 35. Meyer NP, Roelink H. The amino-terminal region of Gli3 antagonizes the Shh response and acts in dorsal-ventral fate specification in the developing spinal cord. Dev Biol 2003; 257(2):343-355. 36. Bose J, Grotewold L, Ruther U. Pallister-hall syndrome phenotype in mice mutant for Gli3. Hum Mol Genet 2002; 11:1129-1135. 37. Gulacsi A, Lillien L. Sonic Hedgehog and bone morphogenetic protein regulate interneuron development from dorsal telencephalic progenitors in vitro. J Neurosci 2003; 23(30):9862-9872. 38. Anderson RM, Lawrence AR, Stottman RW et al. Chordin and noggin promote organizing centers of forebrain development in the mouse. Development 2002; 129:4975-4987. 39. Houart C, Caneparo L, Heisenberg CP et al. Establishment of the telencephalon during gastrulation by local antagonism of Wnt signaling. Neuron 2002; 35:255-265.

CHAPTER 15

Regulation of Early Events in Cell Cycle Progression by Hedgehog Signaling in CNS Development and Tumorigenesis Anna Marie Kenney and David H. Rowitch* Abstract

H

edgehog signaling is essential for proliferation of neural precursor populations in the developing central nervous system (CNS) and is etiologic in cerebellar brain tumors. Here we will contrast general strategies of cell cycle regulation by growth factors in the developing CNS with the emerging concept of a noncanonical Hedgehog "proliferative pathway" as suggested by published studies. Mechanisms utilized by Sonic hedgehog signaling to promote cell cycle progression in CNS progenitors during development and in adult stem cell populations also contribute to CNS tumorigenesis.

Mitogenic Signaling in the Developing CNS A fundamental problem in development is how to precisely regulate the size of neural precursor pools so as to generate sufficient mature progeny to sculpt a central nervous system (CNS) complex enough to handle the organism's needs for motion, sensation, reaction, and perhaps learning and thought. Some primitive organisms, such as C. etegans, restrict the actual number of cells generated, with each cell having a defined role in the developing and mature organism.^ Vertebrates have evolved a different mechanism, in which neural precursors are produced in vast excess with the unnecessary cells undergoing subsequent apoptosis. Rather than being restricted in overall number, generation of neurons is tighdy temporally restricted. Thus, large populations of neuronal precursors are supplied at specific places and times, such that they have the opportunity to mature and form synapses with the appropriate targets. Superfluous neurons—those failing to make successful synaptic connections—die. During later refinement of CNS circuitry, these synapses may be eliminated or remodeled, but the initial selection of surviving cells is based on strength of synapse formation. How pools of CNS progenitors respond to temporal and spatial cues to undergo high levels of proliferation is only now beginning to be understood. As detailed below, localized CNS precursor pool expansion is likely to occur as a result of coinciding (1) microenvironmental mitogenic stimuli, (2) expression of appropriate receptors and signal transduction machinery, and (3) intrinsic cell cycle regulators, or "clocks". These final two criteria may be defined as "cellular competency" to proliferate. In the following section, we discuss the identity and function of molecular mediators of neuronal progenitor proliferation competency. *Corresponding Author: David H. Rowitch—Department of Pediatric Oncology, Dana-Farber Cancer Institute, 44 Binney Street, Boston, Massachusetts 02115, U.SA. Email: [email protected]


188


Precursors of the Embryonic and Postnatal Brain The development of the vertebrate CNS occurs in four major phases. The first specifies the site of the nervous system and biases the ectoderm toward a neural fate known as neural induction. In the second phase, neurulation occurs followed by an expansion of the stem cells to generate a pool of neural progenitors. At this stage, commitment to a particular cell fate (e.g., neuronal subtype or glial cell) does not preclude further proliferation."^'^ Next, the period of symmetric division progressively switches to asymmetric division, whereby one daughter cell permanendy exits the cell cycle and differentiates into its mature phenotype, while the other daughter cell retains its proliferative capacity. Last, at birth, except for a few specialized regions, including the rostral migratory stream, cerebellar cortex, hippocampal dentate gyrus, and cortically located glial precursor cells, '^ the brain enters a state of replicative quiescence. Unlike some somatic cell types (e.g., hepatocytes and myocytes), neurons are unable to reenter the cell cycle after completing terminal differentiation. Indeed, aberrant re-entry into the cell cycle is associated with apoptotic cell death, a process that has been implicated in the pathology of certain neurodegenerative diseases.^'^ Although this lack of cell cycle flexibility was initially interpreted to mean a lack of turnover of CNS cell structures, it is now understood that pools of undifferentiated "neural stem cells (NSCs)" persist throughout the lifetime of vertebrates.^ Collectively, NSCs provide the basis for neuronal precursor expansion during CNS development, and remain a persistent population in adults to potentially replace lost neurons or generate new cells if required, e.g., during learning processes and following injury. As noted above, neuronal precursors undergo both symmetric and asymmetric types of cell division. The determinants driving symmetric or asymmetric division in neural stem cells are not well understood, particularly in higher-order species such as mammals. Unequal inheritance of cell-cycle regulatory components and signal transduction machinery is likely to play as a role. For example, the heritable distribution of Notch protein^ and the Notch modulator, numb,^ have been implicated in determining whether daughter cells of asymmetric division retain their proliferative capacity or take on a differentiated neuronal cell fate. Since commitment of neural precursors to a specific fate does not preclude continued proliferation, basic mechanisms of cell cycle progression are likely to be similar between neural progenitors with a defined fate and uncommitted precursors. Therefore, general lessons learned from studying a committed population, such as cerebellar granule neuron precursors, may be extended to other types of neural stem cells. Here, we will address the broad question of how undifferentiated neuronal progenitors initially respond to mitotic signals and shift from a quiescent phase into the first and second phases of active cycling.

Regulation of the Cell Cycle by Mitogens in CNS Precursors In neural precursor cells, extrinsic and intrinsic signals regulate the balance of proliferation or growth arrest through effects on the cell cycle machinery. The mitotic cell cycle is composed of four phases: one dedicated to synthesis of the genomic DNA (S), one to mitosis (M), and two gap (Gland G2) phases (Fig. 1). The first gap phase, G l , occurs between the end of mitosis and S phase. During early G l , critical decisions are made to commit to another round of cell division, exit the cell cycle permanendy, or transiently exit to a GO phase. Once committed to a new cell cycle, a cell requires sustained stimuli to reach a restriction point, ^^ after which the cell no longer requires mitogenic stimulation to copy its DNA and continue through the cell cycle. The second gap phase, G2, between the end of S phase and the beginning of mitosis, allows cells to repair replication errors and strand breaks made during DNA synthesis and prepare for mitosis. M phase is typically the shortest phase of the cell cycle, and is followed by reentry into G l , or exit into quiescence (GO). 'Immediate Early Genes" Carry Out Activation of the Cell Cycle Apparatus by Mitogenic Signaling In the developing CNS, cell cycle entry is likely to occur as a result of signaling pathway activation by extracellular factors. Bone morphogenic proteins (BMPs), ' Notch pathway

Cell Cycle Progression by Hedgehog Signaling

189

Transcription of S~phase regulators

Figure 1. Scheme for cell cycle regulation in CGNP and effects of Shh on G l phase progression through effects of N-myc. The cell cycle is divided into 4 phases: G l , S, G2, and M. Entry into the cell cycle and progress through the earliest phases is mediated by G l cyclins and cdks acting downstream of immediate early genes. Schematic depicting the cell cycle phases and associated proteins regulating passage through each phase. Mitogens induce cell cycle progression by inducing expression of immediate early genes (legs), such as N-myc in the case of Shh signaling. legs up-regulate D-type cyclins which form complexes with cdks 4 and 6. Cyclin D:cdk4/6 promotes progression through the G1 phase of the cell cycle by inactivating the ckis p21 and p27, which causes activation of cyclin E:cdk2 complexes. Phosphorylation of cyclin E:cdk2 substrates is necessary for entry into and progression through S-phase. CyclinD:cdk4/6 also phosphorylates Rb, thus releasing E2F transcription factors from Rb-mediated inhibition. E2F proteins regulate expression of genes required for S-phase, including some involved in DNA replication.Rb phosphorylation releases E2F transcription factors from inhibition. Activity of cyclin A:cdk2 regulates progression through S-phase into G2, when cyclin A may also associate with cdkl. G2 is marked by an increase in levels and activity of mitotic cyclins, including cyclin B, in complex with cdkl. ligands, neurotrophins,^^ epidermal growth factor (EGF) and fibroblast growth factor (FGF) family members,^ Wnts,^^ and Sonic hedgehog^^ are among the many factors shown to be important for regulating neuronal precursor proliferation. T h e capacity of these factors to promote proliferation is often cell-type specific. For example, neurotrophins p r o m o t e proliferation of cortical progenitors,^^ whereas activation of the Sonic hedgehog pathway is powerfully mitogenic for precursor cells in the cerebellum, ^^'"^^ retina,^'^ and forebrain.^^'"^ Recently, adult hippocampal neural stem cells^ and post-natal telencephalic progenitors have also been shown to divide in response to Sonic hedgehog signaling, demonstrating a mitogenic role for hedgehog signaling throughout life.

190


;^PDGFjflJ

/ % nn

Mitogens

-^Shh.

"l

RTK

G P C R - " ^ > _ ^ / ,

BE)

Ptch

Smoi

Ce// cycle regulators?

Protein translation

i 1 Immediate early genes

Cyclin degradation

Cyclin repression

Figure 2. Many well-characterized mitogens induce cell cycle progression by activating receptor tyrosine kinases (RTKs). Schematic showing seleaed examples of how aaivation of RTKs promotes proliferation through signaling cascades regulating protein synthesis, proteolysis, and gene transcription. Many growth factors utilizing nonRTK receptors, such as G-protein coupled receptors (GPCRs), also promote activity of pathways downstream of Ras. Indeed, some GPCRs direcdy activate the PDGF or EGF RTKs through intracellular scaffolding mechanisms. The hedgehog pathways proliferative effects are less clear. Shh signaling leads to nuclear localization of Gli transcription factors, which might activate or repress gene expression for cell cycle regulatory components. In contrast, activation of N-myc expression by Shh (Fig. 1 and text) provides a direct connection to the cell cycle apparatus via a noncanonical pathway. Many well-understood mitogens access the cell cycle through classically defined receptor tyrosine kinase-activated (RTK) pathways.^^'*^^ When ligand binding activates such RTKs, a cascade of intracellular phosphorylation events results in activation of pathwâys regulating protein synthesis, and inhibition of pathways promoting cell cycle exit. The protein-synthesis dependent induction of so-called "immediate early genes (legs)" also occurs in response to mitogens (Fig. 2). The name "immediate early genes" reflects the fact that mRNAs for such factors were initially characterized by rapid transcription within minutes of serum or specific mitogen exposure in synchronized cells, without requiring new protein synthesis. ' T r a n scription of these genes occurs directly as a result of activation and nuclear localization of cytoplasmic proteins responsive to specific mitogens, including those activating pathways through mechanisms other than RTK binding. Intracellular cytoplasmic sensors, for example Smad proteins downstream of TGF-p, or p-catenin downstream of wnt,^"^ are latent transcriptional regulators, whose ability to activate immediate early gene transcription depends on interaction with nuclear-localized entities, such as chromatin remodeling complexes. The mechanisms through which RTK-activating mitogens such as platelet-derived growth factor (PDGF) promote proliferation have been well studied. By contrast, how the Shh pathway regulates the cell cycle is less clear (Fig. 2).


191

Many of the legs induced by mitogens, such as members of the fos, jun, and myc families, are themselves transcription factors, whose function is to activate genes controlling entry into Gl of the cell cycle. Other known ieg targets include metabolic regidators, which will increase levels of protein synthesis required for proliferation.^^ As described below, N-myc is an ieg with a critical role in Shh-mediated cerebellar precursor proliferation. G l Phase Progression Is Mediated by D-Type Cyclins Among the earliest responders to ieg activity are D-type cyclins. Indeed, cyclins D l and D2 are known to be transcriptional targets of myc. ^'^^ Because their induction is dependent upon activity of upstream regulators, D-type cyclin upregulation is referred to as a "delayed early response."^ In quiescent GO-phase cultured cells, D-type cyclin expression is rapidly elevated in response to mitogenic stimulation, and is typically maintained through the length of the cell cycle, or until growth factors are removed. In primary cerebellar granule cell cultures, protein synthesis-dependent up-regulation of cyclins D l and D2 can be observed within one to three hours of Shh stimulation,^ '^^ with levels dropping when Shh is removed. The timing of cyclin D l up-regulation in primary cerebellar cultures is consistent with the kinetics of the delayed early response in other cell types to mitogenic stimuli. For example, increased levels of cyclin D l can be detected in serum-stimulated keratinocytes after 1 hour.^ These observations indicate that hedgehog-stimulated proliferation of neuronal precursors utilizes mechanisms conserved with other mitogenic pathwavs. D-cyclins are nuclear proteins,"^ which function in several ways to promote progression from GO into and through early phases of G l . In complex with cyclin dependent kinases (cdks) 4/6, D-cyclins sequester the cdk inhibitors (ckis) p21 and p27, which can block S-phase entry though their inhibition of cyclin E/cdk2 activity. ^' ^ Thus, cyclin D:cdk4/6 activity is necessary for cyclin E activation.^^'^ A critical role for D-type cyclin/cdk complexes is to initiate hyperphosphorylation and inactivation of the retinoblastoma protein (pRb), a tumor-suppressor ^ (Fig. 1). The D-type cyclin family comprises three proteins, whose expression and activity during development and in response to mitogens is highly cell-type specific. Studies of mice lacking members of this family have shown that D-type cyclins play an important role in development of many tissues, including the CNS. Specifically, cyclin D l is required for normal retina and mammarv gland development; ' '^ cyclin D2 is important for development of the ovaries and testes; and D3 plays an important role in the lymphoid compartment. Consistent with its important role in mammary development, de-regulated activity of cyclin D l is observed in breast cancers. ^^ Cyclin Dl-mutant animals display neurological abnormalities, although their cerebella appear phenotypically normal. Loss of cyclin D2 results in impaired survival of several cerebellar cell types. Interestingly, combined loss of cyclins D l and D2 causes markedly perturbed cerebellar development, most likely attributed to defective granule cell precursor proliferation.^^ Cyclin D3 ablation has no effect on brain development. Cyclin E Plays a Role in Late G l and the Transition to S-Phase Increased expression of cyclin E is also observed during G l . Like cyclin D, cyclin E is a nuclear protein. However, unlike D-cyclins, cyclin E is rapidly degraded upon entry into S-phase. '^ Activity of cyclin E and its partner, cdk2, occurs later in G l , and functions to facilitate entry and passage through S-phase. Cyclin E activation does not occur as a direct response to mitogens, but is instead regulated by cyclin D-mediated inactivation of the cdk 2 inhibitors p21 and p27 (Fig. 1). Interestingly, mice lacking p27 show marked cerebellar hyperplasia and increased proliferative response to Shh, demonstrating an important role for cyclin E activity in the hedgehog proliferative pathway.^^ In vitro, granule cell precursors treated with Shh show increased levels of cyclin E. Like D-cyclins, cyclin E/cdk 2 can phosphorylate pRb. Replacing cyclin D l with cyclin E results in rescue of cyclin D knockout phenotype, indicating

192


that a major function of cyclin D is its activation of cyclin E. In addition to pRb, cyclin E substrates include numerous proteins involved with S-phase progression,^ '^^ including E2F transcription factors,^ which regulate expression of molecules initiating DNA replication. Over-expressing cyclin E in fibroblasts accelerates progression through G l , likely due to early activation of S-phase regulators.^^ There are two known cyclin E family members, cyclins El and E2. In mice, ablation of either family member alone is not lethal—although cyclin E2 knockout males exhibit reduced fertility.^^ Mutation of both cyclin El and E2 causes placental defects, and live pups are not born. When experiments were performed such that double mutant animals could develop in utero with wild-type placental tissue, they died on the first day of life. Analyses of embryonic development revealed cardiovascular abnormalities. Since the animals died before the onset of CGNP expansion, development of the cerebellum could not be analyzed. Analysis of fibroblasts derived from cyclin E mutant embryos revealed an inability to reenter the cell cycle upon mitogenic stimulation. Given that cerebellar granule neurons cannot be induced to divide in response to Shh after leaving the cell cycle, it is tempting to speculate that this may be due to an irreversible "shut-down" of cyclin E. Indeed, such a "shut down" could be accomplished by strongly increasing levels of ckis, such as p21 or p27. High expression of p27 is observed in post-mitotic granule cells.^^ The Transition from G l to S-Phase Requires Rb Inactivation Both cyclin D and cyclin E in complex with their cdk partners can phosphorylate the tumor suppressor retinoblastoma gene product, pRb. Rb is phosphorylated at several sites, and this phosphorylation is sustained throughout the cell cycle until the cell exits mitosis. In G l , phosphorylation of Rb by cyclin D is likely to be more relevant than phosphorylation by cyclinE:cdk2, as cyclin D is dispensable for proliferation of Rb-null cells, whereas cyclin E is not, indicating that cyclin E s other ftinctions are critical for cell cycle progression. In its hypophosphorylated form, pRb prevents transition from Gl to S-phase, by interacting with and blocking function of specific proteins. Most well known of pRb substrates are the E2F transcription factors, many of which promote S-phase progression through activation of their target genes (Fig. 1). However, Rb is also involved in down-regulating activity of the Ras pathway, which is critical for survival and proliferation in the CNS. This function of Rb is separable from its E2F-binding role. Moreover, Rb can also function as a corepressor with GDP/ cut/cux proteins, blocking activation of cell cycle regulated histone promoters. CDP/cut/cux family members are expressed in the CNS and are implicated in cell fate specification and differentiation, ' ^ consistent with a role for their cooperation with pRb in this process. Recendy, work in osteosarcoma cell lines has shown that phosphorylation of Rb is also important for entering G l from GO. Cyclin C/cdk3 mediates this phosphorylation. It is not known whether a similar process occurs in proliferating neural precursors. However, it is interesting to speculate that suppression of cyclin C/cdk3 activity may be part of the mechanism preventing differentiated neurons from reentering the cell cycle. Rb is a member of the so-called "pocket protein" family, which also includes pi07 and pi30. The "pocket" refers to a highly conserved protein-protein interaction domain, through which binding to E2F factors is mediated. These proteins show specific developmental expression patterns and tissue distributions. ^ P107 and pi30 are more highly related to each other than to pRb, and ablation of either of these proteins in the mouse causes no apparent phenotype during development or adulthood. However, combined mutation of pi07 and pi30 results in defective bone formation and perinatal death due to respiratory defects. The bone formation defect was attributed to excessive proliferation of chondrocytes, indicating that like pRb, pi07 and pi30 have anti-proliferative functions. Neither pi07 nor pi30 have been implicated in cancer. In contrast, mutation of Rb in humans results in retinoblastoma, as well as several other types of tumors, including brain tumors.^^ Several groups have generated mice with homozygous inactivation of Rb.^^^ These animals die by day 15 of gestation, with defects in many


193

tissues, including the hematopoetic compartment. The nervous system is also highly affected: widespread neuronal cell death was observed. A close analyses of Rb-deficient mouse nervous system development showed that many of the dead cells in the nervous system had recently attempted to enter the cell cycle, and that neuronal cell death occurred during a time at which, during normal development, Rb is hypophosphorylated in the nervous system. These findings indicate that Rb's cell cycle inhibitory fiinction promotes survival of post-mi to tic neurons. Cell death in the absence of Rb occurs through pathways, which involve p53, and others which do not. Interestingly, combining loss of p53 with loss of Rb in the developing cerebellum results in medulloblastomas of granule neuron precursor origin, suggesting that the p53 and pRb tumor suppressor proteins may cooperate in promoting normal cell cycle exit in the developing cerebellum.

SHH Signaling Is Required for Growth of Cerebellar Granule Neuron Precursors and Is Etiologic in Medulloblastoma The cerebellum has pivotal roles in coordination of posture and locomotion, head and eye movements and a wide range of routine and skilled motor activities. Several lines of evidence further indicate cerebellar functions in learning, cognition and language. The stereotypical laminar organization of the cerebellar cortex—^with its limited number of input and output lines, and only a half dozen major classes of neurons—has aided efforts to trace successive stages of cerebellar ontogeny and molecular mechanisms involved in its development (Fig. 3). Purkinje cells and the cells of the cerebellar nuclei originate from precursors of the rhombic lip and caudal midbrain. The precursors for granule cells are generated in rhombomere 1 of the embryonic hindbrain and migrate dorsally to form the outer layer of the cerebellum, or external granule layer (EGL). Certain transcription factors including Math-1 have been demonstrated to be essential for early granule cell development. Proliferation of granule cells is largely postnatal.'^^ Proliferating granule cell precursors are readily detected in the cerebellum by virtue of their position in the EGLa compartment. Post-mitotic cells initially become NeuroDl-positiyc and reside in the EGLb compartment. They then migrate to the IGL, comprised of differentiated granule cells that label with the mature neuronal marker, NeuN, and the granule cell marker, Zic. In humans, granule cell differentiation is completed by approximately 18 months of age. The equivalent process in the rodent cerebellum takes place within 1 month of birth.^^ SHH Signaling Is Essential for CGNP Proliferation During mammalian central nervous system (CNS) development, multipotent precursor cells undergo division, cell fate specification, and maturation in response to extrinsic cues. The secreted signaling molecule Sonic hedgehog (Shh) is essential for development of organizing structures at the ventral midline (e.g., floorplate) and the specification of neurons and glia. In addition, recent evidence has indicated that Shh regulates the proliferation of granule neuron precursors in the cerebellum.^^'^^ Proliferative effects associated with Hedgehog pathway activation have also been described in the developing neural tube'^^"^^ and retina. The active Shh signal is produced by autoprocessing and cholesterol modification, and binds to a receptor complex composed of at least two transmembrane proteins. Patched and Smoothened.'^ '^^ Shh binding to Patched is thought to relieve Patched-mediated inhibition of Smoothened activity, resulting in the activation of transcriptional targets by members of the Git family.^ Smoothened belongs to the family of serpentine G-protein coupled receptors (GPCRs). Shh signaling can be inhibited experimentally by increasing cyclic AMP levels or protein kinase A (PKA) activity. Developmental effects of Shh can be mimicked in vivo by expression of pertussis toxin^^ or dominant negative PKA,^^ suggesting that an inhibitory G protein (Gai) may be the target of Smoothened. However, a specific heterotrimeric G-protein downstream of Smoothened has yet to be identified,^^ and endogenous cyclic AMP levels do not respond to Hedgehog pathway activation.^^ Conserved components of the Hedgehog

194


Figure 3. Structure of P7 mouse cerebellum and expression of mRNA transcripts for N-myc the EGL layer in situ. The layered architecture of the cerebellum is apparent in the cartoon of a parasagittal section at low power. A) High power view showing layers of the developing cerebellum. The EGLa comprises proliferating Math-1 -positive granule cells. Commensurate with maturation to NeuroD 1 -positive nonproliferating cells, they migrate to the EGLb layer. Subsequent stages of maturation are recognized by the markers Zic and NeuN, and further migration to the IGL. The Purkinje neurons are the source of Shh proteins, which are exported to the EGLa compartment through a poorly understood mechanism. The cerebellar white matter (CW) layer is indicated. B) N-wyr is expressed in proliferating precursors of the developing cerebellum. N-wj/f expression is observed during the proliferative phase of cerebellar development at PN7, but is undetectable in the FN 15 or adult (not shown) mouse cerebellum. Low power views show the distribution of N-wjf mRNA transcripts primarily in the EGLa region. signaling pathway include Fused and Suppressor of Fused. ^"^'^^ These proteins are thought to retain the Shh-activated transcription factors Gii2 and Gli3 (orthologues of Drosophila ct) in the cytoplasm via Costal2-mediated interactions with microtubules.^ ' Until recently, the Shh signal transduction pathway was not known to share common targets with any known mitogenic intracellular signaling pathways. Indeed, it has been proposed that proliferative effects of Shh on retinal precursors are indirect, perhaps involving synthesis of


195

a secondary mitogen. ^ However, a secondary mitogen need not be invoked, as transactivation of receptor tyrosine kinase (RTK) pathways by GPCRs is well described.^^'^ Upon phosphorylation by GPCR kinases (GRKs), GPCRs can utilize the intracellular domains of RTKs as scaffolds for stimulating activation of the mitogen-activated protein kinase (MAPK)/ extracellular-signal regulated kinase (ERK) pathway, a target of many extracellular mitogenic stimuli."^^ Independent of RTK scaffolding, some GPCRs can activate the MAPK pathway through PI-3 kinase activation.^^ This mechanism involves the p/y subunits of heterotrimeric G-proteins. Given the resemblance of Smo to a GPCR, and recent evidence that Smo activates and endogenous Ga, it is possible that Shh proliferative signaling through Smoothened may also regulate MAPK in neuronal precursors. However, it is unlikely that this mechanism plays a role in Shh-mediated cell division, as inhibiting MAPK activity does not affect Shh-mediated proliferative effects on cerebellar granule neuron precursors.^^ SHH is required for granule cell precursor proliferation during cerebellar development. Purkinje cells arborize towards the EGL and are considered the relevant source of SHH signal for granule cells in the EGL (Fig. 1). The Hedgehog pathway is highly conserved (Fig. 2). Genetic and biochemical characterization in Drosophila has yielded important insights into its organization in higher vertebrates. ' This has led to the proposal that SHH relieves Patched-mediated inhibition of Smoothened activity, resulting in the activation of transcriptional targets by members of the Gli family.^ Inappropriate Activation of H H Signaling Results in Brain and Other Tumors Previous work in the fields of developmental biology and human genetics has revealed that certain mechanisms essential during formation of the embryonic CNS can also become inappropriately activated at later stages resulting in tumors of the cerebellum. Activating mutations of the Wnt and Hedgehog signal transduction pathways, in particular, have been associated with medulloblastoma, a cerebellar tumor most commonly affecting children. Loss-of-fiinction mutations in human PATCHED (PTC) are associated with activation of the Hedgehog signal transduction pathway and promotion of a neoplastic state. Three to five percent of children with Gorlins syndrome, caused by inherited mutations of PTC, develop medulloblastoma.^ Inactivating mutations of PTC have also been found in sporadically occurring medulloblastoma; mice heterozygous for targeted mutations of Patched, in which SHH targets are potentially upregulated, develop cerebellar tumors at a low incidence (10-15%).^^^ Interestingly, combination of the PTC+/- allele with loss of the tumor suppressor p53 results in a dramatic increase, with tumor incidence approaching 100%.^^ Althoi^h the combination of PTC+/mutation and p53-l- is not found together in human tumors, these data suggest that other "second hit" mutations might facilitate tumor progression in combination with deregulated SHH signaling (Fig. 2). Indeed, experimental mutations in Rb and p53 also promote medulloblastoma formation in mice.^^^ Further work is needed to identify the authentic collaborative mutations that may occur within human medulloblastoma. Pomeroy et al have recently shown by oligonucleotide microarrays that the desmoplastic sub-type of medulloblastoma is specifically associated with activation of Hedgehog gene targets. This sub-type comprises --20% of all medulloblastoma cases in the US, and is thought to result from loss-of-function mutations of the Hedgehog pathway suppressors, PTC and SUFU}^^^\v^ mechanisms connecting Hedgehog signal transduction to molecular regulators of the cell cycle during development and tumorigenesis are beginning to be understood through biochemical analyses of Shh proliferative effects on responsive neuronal precursor populations, as well as phenotypic analysis of mouse genetic models. Understanding Sonic Hedgehog Pathway Interactions with the Cell Cycle Apparatus The links between the Shh pathway and the cell cycle apparatus have been less clear than with many other mitogens so far studied. Downstream effectors of cell cycle progression have been highly conserved through evolution, making it likely that the Shh pathway intersects with

196


similar molecular sensors to manifest its mitogenic role in competent neuronal progenitors. However, the studies of canonical Shh pathway, as defined by genetic analyses, have failed to clearly indicate this mechanism, necessitating the use of biochemical approaches. Here we will examine the candidate mediators of Shh effects on regulation of Gl progression, initially identified in primary mouse CGNP cultures and subsequendy validated in vivo. During the phase of granule cell expansion in the mouse, it is relatively straightforward to prepare primary cultures consisting of approximately 90% pure CGNPs. With additional purification, cultures up to 98% pure can be obtained.^^^ CGNPs can be maintained in a proliferative state in vitro in the presence of Shh for a week. It is important to note that the presence of Shh does not constitute an absolute block to proliferation: even with Shh treatment, many CGNPs exit the cell cycle and go on to differentiate,^^ most likely due to intrinsic cell cycle clock mechanisms. However, sufficient numbers of CGNPs proliferate in vitro in response to Shh to make these cultures ideal for performing biochemical analyses of cell cycle regulatory events, modeling effects of signal transduction pathway activation on proliferation, and ectopic expression studies using retroviral infection. By taking advantage of these easily accessible, nearly homogeneous, and highly manipulatable cells, we and others have been able to make great strides in defining mediators of Shh proliferative effects on neuronal precursor cells.

1$ There a Role for GLI Activity in Vertebrate HH Proliferative Signaling? Signaling through the Hedgehog pathway results in activation of Gli zinc-finger transcription factors, ^^ which are cytoplasmically localized until activated. The topic of Gli signaling is covered in several other chapters and will only be touched on briefly here. GUI behaves as a transcriptional activator, whereas Gli3 generally functions as a repressor of hedgehog signaling.^ ^^' ^ Although Gli2 has been shown to have activator and repressor properties distinct from Glil}^^ GUI can functionally compensate for Gli2 during CNS development.^^^ It has been suggested that all three are retained in the cytoplasm, and processed or relocalized in response to hedgehog signaling. ^^^'^^^'^^ Gli proteins are implicated in regulation of neural proliferation, as GLIl was originally identified as a locus amplified in a low percentage of human gliomas. Over-expression of Gli is sufficient to cause skin tumors and neural hyperplasia,^ as well as granule cell precursor proliferation, and transcriptional effects of GLI have been observed on D-type cyclins by gene expression profiling of a transformed rat kidney cell line.^^^ Indeed, these experiments suggest that GUI can directly upregulate cyclin D2. Moreover, Oliver et al have shown that retroviral expression of GUI is sufficient to maintain high levels of CGNP proliferation in vitro, in the absence of Sonic hedgehog, suggesting that GUI can regulate components of the cell cycle regulatory apparatus. However, GUI function is not required during cerebellar growth^ ^^ or tumorigenesis demonstrating that regulation of D-type cyclins and cerebellar growth is not G/zV-dependent in vivo. These experiments do not rule out an important role for GU2, which is expressed in a similar pattern to GUI and may compensate for loss of GUI function. It is interesting to note, however, that proliferation and cyclinDl levels are normal to increased in spinal cord progenitors of compound mutant animals lacking all Gli fiinction.^^^ In vitro, Gli2 is less potent than GUI in its ability to promote CGNP proliferation.^^ Thus, the role of the canonical Shh signaling pathway in mediating Shh mitogenic effects remains controversial, and is likely to be a complicated question to unravel given the combination of activator and repressor functions carried out by Gli proteins. It is also possible that Shh signaling through Ptc and Smo regulates cell cycle progression through novel mechanisms. Indeed, as described below, recent work raises the possibility that Shh access to the cell cycle via the proto-oncogene N-myc may not involve Gli proteins.


197

N-MYC as a Key Component of HH Signaling Identification of Shh signaling intermediates tliat fiinction in cell cycle regulation is key to understanding the role of this pathway in CNS proliferation during development and tumorigenesis. We and others have shown that treatment of CGNP cultures with Shh results in rapid upregulation of D-type cyclin genes, and that regulatory effects on Q\ cyclins require protein synthesis.^^ Like classical mitogens, which promote proliferation through upregulation of immediate early genes, Shh direcdy induces the proto-oncogene N-myc in primary cultures of CGNPs. ]>\-myc is a member of the myc family, which also includes c-myc and h-myc. Activity of myc proteins is known to promote proliferation and/or transformation in many cell types.^ ^ Complete loss of myc activity results in severely compromised cell cycle regulation. ^^^ When expression of myc family members was examined by in situ hybridization, only ^-myc was expressed in the proliferative zone of the cerebellum, suggesting a potential role for N-myc in CGNP proliferation in vivo (Fig. 3). N-myc mRNA was detected in the same region of the developing cerebellum as cyclin D l , and Glil, demonstrating activation of the Shh pathway in proliferating CGNPs in vivo (Fig. 3). Moreover, N-myc expression was only detectable during the time of cerebellar expansion, being completely downregulated by PI5 (Fig. 3). We observed ^-myc upregulation in CGNP cultures and the dorsal embryonic spinal cord exposed to Shh, suggesting that signaling via N-wj/c may be a general feature of Shh-induced proliferation in the CNS. However, in contrast to known general transcriptional targets of Shh (e.g., Gli genes), we note that Shh signaling effects on ^-myc regulation are unique to immature, proliferating precursor cells. Thus, cellular competence is evidendy a critical determinant of the transcriptional response of N-wj/c to Shh. Increased levels of AZ-AfFCmRNA have been reported in certain cases of human meduUoblastoma.^"^^ More recently, AT-AfFC upregulation has been observed in the desmoplastic form of meduUoblastoma, the specific subtype of meduUoblastoma associated with pathological activation of the Hedgehog pathway. In contrast, N-MYC expression was not elevated in the majority of other (nondesmoplastic) meduUoblastoma types analyzed. Consistent with findings in humans, we observed dramatic ^-myc upregulation in medulloblastomas from Ptc heterozygotes.^^^ Together, these findings are strong evidence that similar mechanisms operate downstream of Hedgehog signaling during central nervous system development and tumorigenesis. Our results further suggest that N-MYC expression in desmoplastic meduUoblastoma is a direct consequence of Hedgehog pathway activation.

Evidence That N-myc Activity Is Required for the Full Shh Proliferative Response in CGNPs To establish a role for N-myc in the Shh proliferative pathway in developing cerebellar precursors, it will be necessary to examine cell cycle progression in the absence o^^-myc function. However, mice homozygous for ^-myc null or certain hypomorphic alleles die in utero or at birth, ' preventing assessment of ^-myc function during the proliferative post-natal phase of CGNP development. Recendy, Knoepfler et al have used lox-cre technology to conditionally target murine ^-myc in the developing CNS, resulting in severe hypoplasia of cerebellum and defects in granide cell precursor proliferation. ^^^ These findings indicate that N-wj/r function is necessary for cerebellar development. Yet, they leave open the question of whether N-myc activity is required downstream of Shh signaling during the post-natal phase of granule cell expansion. In vitro studies suggest this is likely to be the case, as dominant negative inhibition of N-myc results in drastically reduced cerebellar precursor proliferation, '^^^ and ere retrovirus infection of CGNPs cultured from N-myc .mice causes near complete cessation of Shh-inducible proliferation (A.M. Kenney, D. Rowitch, P. Knoeptler and R. Eisenman, unpublished). Cre-mediated ablation of N-myc in early post-natal granule cell precursors is likely to shed light on this issue in vivo.


198

Myc box 1 Phosphorylatlon/degradation ISO S54

\/

•4 Myc box 2 TRRAP interaction

LII—i

• 1 -J

z

1 1

1

bHLH/LZ Max dimerization DNA binding

Figure 4. N-myc protein structure contains regions highly conserved among myc family members. A schematic illustrating functional domains ofN-myc, which migrates at approximately 50 KD on SDS-PAGE. The amino terminal comprises the transactivation domain. This portion of the protein includes the myc box one, which contains phosphorylation sites regulating mvc stability. C-myc box one has been proposed to interact with Binl, a protein involved in differentiation.^^ Binl interaction with N-myc has not been investigated. Myc box 2 mediates interactions with TRRAP, a protein involved with chromatin remodeling and whose interaction with myc is required for activation of certain target genes. ^^^ TRRAP interaction promotes histone acetylation. This region of N-myc is required for promoting CGNP proliferation. The carboxyl terminus of myc proteins contains basic helix-loop-helix and leucine zipper domains responsible for dimerization to Max, and DNA binding. In studies of myc biology, over-expression of Mad family members is an established method for inhibiting myc functions in proliferation, transformation and apoptosis. These studies indicate that the repressive activity of M a d family members is specific, and M a d proteins are unlikely to behave as general transcriptional repressors; indeed, Mxi over-expression cannot inhibit transformation by E l a, another potent oncoprotein. ^ ^ Mxi over-expression was shown to abrogate the mitogenic effects of Shh in vitro, raising the possibility that Mad family members may regulate C G N P cell cycle exit during cerebellar development in vivo. In keeping with this, expression profiling studies indicate up-regulation of several A / ^ family members at F N 7, coinciding with the major wave of C G N P diflFerentiation ( Q . Zhao, A. Kho, I. Kohane and D . Rowitch, unpublished observations). Although the levels of C G N P proliferation in vitro are reduced in the presence of dominant negative N-myc,^^'^^^ absolute proliferation arrest is not observed. This is not surprising, since some residual N - m y c activity is likely to remain. W h e n myc activity is reduced, but not eliminated, myc-dependent processes can still occur. ^^^ Indeed, cultured fibroblasts lacking any myc family member remain capable of proliferation albeit with an extended cell cycle length. ^"^^ T h u s , abrogation of N - m y c activity might not necessarily lead to cell cycle exit within the time frame of the in vitro studies performed. Taken together, the finding of decreased C G N P proliferation with specific N - m y c antagonists in vitro, coupled with data that demonstrate essential functions for N-myc during cerebellar development in vivo, strongly suggest that N-myc fiinctions as an integral component of a Shh regulated pathway during proliferation of granule neuron precursors. M e c h a n i s m s U n d e r l y i n g Cell Cycle Regulation b y N - m y c i n Proliferating Neuronal Precursors Myc proteins are characterized by specific functional domains (Fig. 4), including a carboxyl terminus D N A binding domain and a helix-loop-helix/leucine zipper domain, which mediates dimerization with mycs obligate partner. Max (Myn in the mouse)^^^).^'^^ Two highly conserved amino acid regions, termed Myc box 1 and Myc box 2, reside in the amino terminal transactivation domain. Myc box 1 contains two phosphorylation sites with important roles in


199

regulating myc stability. Myc box 2 associates with TRRAP protein, which functions to recruit histone acetyltransferase to myc's transactivation domain. ^^^ This interaction is required for the transforming potential of c-myc and N-myc. The Myc box 2 domain also appears to be required for N-myc's proliferative function in CGNPs, as over-expression of an ^-myc mutant lacking Myc box 2 cannot rescue CGNP proliferation in the absence of Shh signal, and interferes with CGNP proliferation in the presence of Shh. It is likely that N-myc activates yet to be identified genes involved in cell cycle regulation. Recently, Id2 was shown to be an N-myc target in neuroblastoma.^ ^ Some c-myc targets include genes such as cyclinD^^ and cdc25, ^ ^ which control cell cycle progression. Metabolic regulators that enhance cell growth may also be targets of c-myc. Finally, c-myc has been implicated in repression of genes that promote cell cycle exit and differentiation. Given the high level of relatedness between N-myc and c-myc protein structure, and the observation that N-myc can rescue many of the defects in c-myc null mutant mice, ^ it is possible that they have similar transcriptional targets. We observed increased levels of cyclin Dl in N-wyr-infected cells despite the presence of cyclopamine. ^-myc is a direct transcriptional target of Shh signaling, whereas cyclin Dl mRNA upregulation was inhibited by cycloheximide.^^ N-myc could be a candidate regulator o^ cyclin Dl expression. Conversely, N-myc may repress or inhibit expression of proteins involved in cell cycle exit. Indeed, mice conditionally mutant for N-myc in the CNS show high levels of cki expression in the brain, consistent with premature cell cycle exit. In summary, several studies indicate that N-wjc is a direct downstream target of the canonical Shh signaling pathway in proliferating CGNPs. ^-myc expression was also found in medulloblastomas o^Ptc mutant mice, and in humans with Shh-pathway associated medulloblastomas. N-myc function is a key component of a Shh proliferative pathway essential for normal expansion of CGNP populations. These studies identify a direct connection between the Shh signaling pathway and a cell-intrinsic regulator of the cell cycle apparatus in primary neuronal cells and cancer cells. Proximal events in the progression from Smoothened activation to upregulation o^^-myc expression remain to be determined. Identification of N-myc transcriptional targets will be critical for understanding of Shh-mediated cell cycle progression in CNS development. Moreover, because cerebellar granule cells are postulated to be the cell-of-origin for meduUoblastoma, effective means to inhibit N-myc activity might provide new approaches to control the growth of cerebellar tumors. In the following section, we will cover work on the overall regulation of N-myc activity levels in CGNP, which involves combinatorial interactions between Shh and PI-3kinase signaling. Regulating N-myc Protein: Post-Translation Modification and Turnover Identification of the intracellular events that integrate effects of divergent signaling pathways is critical for a comprehensive understanding of growth control in both normal and neoplastic neuronal progenitors. To better understand interactions between the Shh and other signaling pathways during neuronal precursor proliferation, we have focused on determinants of N-myc protein turnover and cell cycle progression in primary CGNP cultures. C-myc, L-myc and N-myc all feature residues in the highly conserved amino-terminal myc box 1 (MBl) domain that could function as phospho-acceptor sites. Mutation of C-MYC amino-terminal phosphorylation sites in humans is associated with Burkitt's and other aggressive lymphomas, '^ suggesting that disruption of the regulatory role played by these sites may contribute to human disease. Previous analysis of c-myc has implicated phosphorylation of amino-terminal sites in c-myc turnover ' using rat embryonic fibroblasts, cell lines or nonmammalian cell systems. In contrast, N-myc phosphorylation had not previously been established. We found that dual phosphorylation on threonine 58 and serine 54 promotes N-myc protein turnover and timely cell cycle exit in CGNP primary cultures. Signaling by Sonic hedgehog appeared to not be directly involved in N-myc phosphorylation, suggesting that N-myc protein turnover is regulated separately from its mRNA induction by Shh in CGNPs. N-myc

200


mutants unable to be phosphorylated at either or both sites are extremely stable, and promote prolonged proliferation in cultured CGNPs. These results suggest that kinases acting at these sites may promote cell cycle exit through their ability to increase N-myc degradation. N-myc threonine 50 is located within a consensus sequence for phosphorylation by glycogen synthase kinase 3P (GSK-3P). Phosphorylation at this site requires a priming phosphorylation event at S54, another characteristic of GSK-3P activity. Moreover, GSK-3P is known to antagonize proliferation in general, by promoting proteolysis of cyclin Dl,^ ^ as well as downregulating hedgehog signaling by inducing proteolysis of Ci, the fly ortholog of Gli.^^^ Indeed, we found that inhibiting GSK-3p activity prevented phosphorylation of T50.^^^ The identity of the kinase acting at S54 remains to be determined, although several candidates for the analogous site (S62) in c-myc have been proposed. These include MAPK, JNK, and members of the cdk2 family. In vitro kinase assays have shown that MAPK is capable of phosphorylating S62. However, modulating activity of the MAPK pathway in cultured cells has no effect on c-myc S62 phosphorylation.^^^ As discussed above, the MAPK pathway appears to play no role in CGNP proliferation, suggesting it may not be involved in N-myc phosphorylation. Indeed, preliminary studies have found that activation or inhibition of this pathway does not affect N-myc phosphorylation or stabilization in CGNPs (DHR and AMK, unpublished obs). The activity of GSK-3|3 is kept in check by the phosphoinositide 3 (PI-3) kinase pathway. IGF 1-mediated activation of the PI-3 kinase pathway is critical for long-term survival of cultured CGNPs,^^^ and under our culture conditions, insulin is present at levels sufficient to activate the IGF receptor. We found that short-term withdrawal of insulin substantially destabilized N-myc in cultured CGNPs, and this de-stabilization could be prevented by substitution of insulin with IGF. IGF signaling is important for CNS development, and increased IGF activity results in cerebellar hyperplasia. ^^^' In addition to enhancing growth by promoting survival, IGF-mediated activation of PI-3 kinase also has positive effects on the cell cycle regulatory apparatus. PI-3 kinase negatively regulates forkhead transcription factors, which can promote cell cycle exit by repressing cyclin D.^^^ The finding that IGF-stimulated PI-3 kinase activity can stabilize N-myc through inhibition of GSK-3P provides additional insight as to the molecular mechanisms underlying the pro-proliferative effects of PI-3 kinase signaling in neuronal precursors. Several extracellular signals have been proposed to modulate proliferative effects of Shh in CGNP cultures, including those that activate RTK, Notch, and chemokine receptors.^ '^^' The precise intracellular mechanisms underlying such interactions are unclear. However, many of the signaling pathways implicated in cerebellar growth control have direct or indirect effects on the PI-3K pathway. For example, even though our work has rided out a direct role for the Shh pathway in phosphorylating N-myc, it is known that several transcriptional targets of Shh include proteins which regulate IGF signaling and PI-3K activity,^^ suggesting an indirect role for Shh signaling in regulating N-myc stability. Control of N-myc expression and turnover provides an example of how regulation of a single intracellular target can integrate activities of divergent signaling pathways in the developing brain (Fig. 5).

Conclusions and Future Prospects Our review has focused on how the Shh signaling pathway promotes entry into and progression through the G1 phase of the cell cycle in proliferating neuronal progenitor cells. In cerebellar granule cells, Shh induces expression of the proto-oncogene N-myc, which in turn activates crucial mediators of Gl progression, including cyclins D l and D2. In keeping with this mechanism, we have proposed that cerebellar granule cell cycle exit is regulated, at least in part, by amino-terminal phosphorylation of N-myc, which promotes its destabilization. N-myc phosphorylation requires activity of GSK-3P; a kinase with several previously defined anti-proliferative and hedgehog-antagonistic roles. GSK-3p activity is negatively regulated by the PI-3K pathway, which is known to be required for cerebellar granule proliferation and survival. Together, the available data indicate that Shh and PI-3 kinase act at distinct steps of


201

Shh

Growth/survival factors (IGF)

H-myc mRNA

PI-3 kinase

1

Priming kinase p T

GSK3 pp •

Target gene regulation

Q ^^S^^Ô

Cell cycle progression

Cell cycle exit

Figure 5. Integration of signaling pathways promotes proliferation during cerebellar development. Schematic illustrating how two pathways with known positive effects on neuronal precursor proliferation can cooperate by convergence on a common target. Shh signaling induces expression of the proto-oncogene transcription factor 'H-myc. N-myc is necessary and sufficient for maintaining CGNP proliferation. Maintenance of N-myc protein levels in vitro is regulated in a Shh-independent manner. We identified a role for the PI-3K pathway in stabilizing N-myc protein. In cultured CGNPs, IGFR signaling activates the PI3 kinase pathway, which inhibits GSK3-mediated phosphorylation of N-myc T50 and its subsequent degradation. The kinase priming N-myc for GSK3 phosphorylation, by acting at S54, remains to be identified. Other growth factors (e.g., SDF-1) or cell-cell interactions (e.g., via integrins) are candidates for PI-3 kinase regulation and inhibition of GSK3 activity in vivo. The combined effect of concerted Shh and PI-3 kinase activity is to promote G1 progression and precisely control the timing of cell cycle exit. A similar theme may be found in medulloblastoma tumorigenesis, as both PI-3K and hedgehog pathway activation are associated with this disease. m R N A transcript upregulation and protein stabilization, respectively, to regulate overall activity and levels of a c o m m o n target, N-myc.

Common Mechanisms in CNS Development and Tumorigenesis: Sonic Hedgehog Signaling as a Therapeutic Target in Medulloblastoma In this review, we have emphasized many parallels between regulation of the cell cycle in developing C G N P and in h u m a n cases of medidloblastoma. W e have recently taken a global approach to this question and compared many thousands of genes upregulated both in desmoplastic and classic variants of medulloblastoma to a spectrum of genes normally expressed during mouse cerebellar development, as assessed by oligonucleotide microarrays. Genomically, h u m a n medulloblastomas were closest to mouse F N 1-10 cerebella; normal h u m a n cerebella to mouse P N 30-60 cerebella. Furthermore, metastatic medulloblastomas were highly associated with mouse P N 5 cerebella, suggesting that a clinically distinct subset of tumors is identifiable by molecular similarity to a precise developmental stage. Together, our findings indicate a global recapitulation of tissue-specific developmental programs in medulloblastoma, and the utility of tumor characterization on the developmental time axis to identify novel aspects of clinical and biological behavior. ^^^ Although fiirther work is required to expand o n these observations, for example by testing roles for individual genes in mouse models of medulloblastoma, it is clear that important parallels between C N S development and tumorigenesis exist at the genetic and mechanistic level.

202


The established participation of Hedgehog signaling in meduUoblastoma/^^'^^^ coupled with the availability of inhibitors specific for Hedgehog-Smoothened signaling^ ^^'^ has raised the exciting possibility that such inhibitors may be useful clinically. The Hedgehog inhibitors jervine and cyclopamine were originally isolated from plants, and synthetic inhibitors of Hedgehog signaling have been developed commercially. Providing preclinical data is forthcoming from mouse studies, it seems likely that such Hedgehog pathway inhibitors will soon be available for pediatric patients with medulloblastoma. Recendy, the Shh pathway has been shown to be mitogenic for NSC populations and required for maintenance of the stem cell niche. ^ This data indicates that a possible adverse side effect of Hedgehog pathway blockade in children could be a depletion CNS neural stem cell pools, of particular concern for long-term effects on neurodevelopment and cognition. Such potential side effects must be weighed against the overall toxicity of currendy used therapies such as chemo- and radiotherapy. In addition to general blockade of Shh signaling, it may be possible to target the downstream mediators of Hedgehog signaling with specific roles in cell cycle regiJation. For example, our work suggests that inhibition of PI-3 kinase activity may be sufficient to destabilize N-myc. Further elucidation of Hedgehog mitogenic effects, including how the Shh pathway functions in context of other pathways, may provide further clues for therapeutic intervention in human tumors. This issue takes on broader significance when one considers that Shh-Smo signaling is implicated in a wide variety of human cancers.^ '

N'tnyc as a Nodal Point for Interacting Pathways Regulating CGNP Proliferation The events we have studied in CGNPs are likely to be recapitulated in human tumors of cerebellar origin. For example, N-myc is up regiJated in Shh pathway-associated medulloblastomas in mice and humans. Moreover, evidence of increased activity of the PI-3K pathway is also observed in human and mouse medulloblastomas, ' as well as other types of brain tumors, such as adult gliomas. Such increased activity, with the consequence of inhibiting GSK-3p, may promote tumor cell growth by stabilizing N-myc and other cell cycle regulatory proteins, including D-type cyclins. These connections suggest several targets for designing specific therapies against medulloblastomas, including the Shh pathway itself, N-myc, and the PI-3K pathway. Cell growth control is a tighdy regulated process. Although much attention has been paid to the importance of the Shh pathway in regulating proliferation of neuronal precursors, it is necessary to keep in mind that Shh effects do not occur in a vacuum, and that proliferation, cell cycle exit, and cell fate outcomes result from finely tuned responses to Shh interactions with other signaling pathways operating in these cells. A major area of future exploration will be devoted to identifying nodes of cooperative or antagonistic interactions of the Shh signaling pathway with other major regulators of proliferation and brain development. N-myc itself exemplifies such a node. We have already described how the PI-3 kinase pathway cooperates with Shh to promote N-myc-mediated proliferation, through stabilizing effects on N-myc s phosphorylation state. However, such stabilization may also be achieved through the action of phosphatases, removing the phosphate groups at T50 and S54, which direct N-myc s degradation. Although phosphatases targeting N-myc have not been identified, PP2A has recently been implicated in regidation of c-myc stability. Microarray studies have identified SETl, an endogenous PP2A inhibitor, as a target of Shh in proliferating cerebellar granule cells. By regulating the expression of phosphatases and their inhibitors, Shh could indirecdy promote continued proliferation, or timely cell cycle exit, facilitated by N-myc stabilization or degradation. Phosphatases such as PPl and PP2A are highlv expressed in the developing and mature brain, as are endogenous inhibitors of their activity. Phosphatase access to their substrates may be affected by the substrate's intracellular localization, conformation, or interactions with other proteins. The prolyl isomerase Pinl is an example of a protein that could prevent or promote phosphatase access to N-myc. Pinl is widely expressed, and has been implicated in


203

positively regulating activity of proteins important in neuronal precursors,^^^ such as tau and (J-catenin. Pinl binds to phosphorylated serine or threonine residues, which are followed by a proline residue. Pinl affects isomerization around the proline, which has been shown to affect the localization, de-phosphorylation, and activity of several cell cycle regulatory proteins, including cyclin D l . A recent study has also suggested a role for Pinl in regulating myc stability, though these results have yet to be validated using in vivo models and endogenously expressed protein. However, Pinl over-expression is implicated in many human cancers, and it is generally understood to have a pro-proliferative role. Interestingly, Pinl is inactivated by PKA, ^^ a classic hedgehog pathway antagonist. In vivo, such inactivation and disruption of Pinl:N-myc interaction could promote cell cycle exit in Shh-stimulated cerebellar precursors. As a Shh-induced target, N-myc may also play a role in the timing of CGNP differentiation. The ability of BMP signaling to promote neuronal differentiation has been well-established. Recendy, Angley et al have demonstrated an important role for BMP4 in the post-natal differentiation of CGNPs.^^ The timing of BMP pathway effector (Smad proteins^^ ) expression overlaps with the phase of Shh-induced CGNP proliferation in vivo. Shh and BMP are also known to oppose eachother in regidating cell cycle progression in forebrain embryonic neural progenitors. Smads cooperate with the transcription factor Mizl, which induces expression of cdk inhibitors, thereby promoting cell cycle exit.^'^^ Interaction of Mizl with myc results in repression of these genes. We have observed Mizl expression in cultured CGNPs (AM Kenney and D Rowitch, unpublished observations. Whether Miz interacts with N-myc has yet to be determined, but given the highly conserved nature of myc proteins, it is possible that N-myc's anti-differentiation properties could arise in part from its ability to titrate levels of Mizl and repress cki expression, until the strength of BMP-mediated Mizl activation dominates, thereby permitting CGNP growth arrest. It is also worth noting that BMP signaling can promote stabilization of PTEN,^"^^ a negative regulator of PI-3 kinase activity, suggesting that BMP signaling could antagonize Shh-mediated proliferation by promoting N-myc destabilization through increased GSK-3P activity. Whether activation of differentiation-promoting pathways such as BMP signaling can be used as an anti-tumor therapy remains to be seen, but it is intriguing to speculate that such strategies could be employed to prevent tumor expansion or promote apoptosis in brain tumors. What is clear from many studies in recent years is that developmental biology provides a vehicle to both ask and answer provocative questions about CNS tumorigenesis and anti-tumor therapy. An exciting possibility is that other CNS cancers, such as glial tumors, will show parallels to their counterpart glial precursors in the developing brain, analogous to the parallels we have discussed here between cerebellar precursors and medulloblastomas.

Acknowledgements A.M.K. is funded by a Distinguished Scientist grant from the Sontag foundation. The studies described were funded by grants to D.H.R. from the NINDS, the Kimmel Foundation and the James S. McDonnell Foundation.

References 1. Kandel E, Schwartz JH, Jessell TM. Principles of Neural Science. 3'*^ ed. New York: Elsevier Science Publishing, 1991. 2. Bally-Cuif L, Hammerschmidt M. Induction and patterning of neuronal development, and its connection to cell cycle control. Curr Opin Neurobiol 2003; 13(1): 16-25. 3. Cremisi F, Philpott A, Ohnuma S. Cell cycle and cell fate interactions in neural development. Curr Opin Neurobiol 2003; 13(l):26-33. 4. Alvarez-Buylla A, Garcia-Verdugo JM, Tramontin AD. A unified hypothesis on the lineage of neural stem cells. Nat Rev Neurosci 2001; 2(4):287-293. 5. Taupin P, Gage FH. Adult neurogenesis and neural stem cells of the central nervous system in mammals. J Neurosci Res 2002; 69(6):745-749. 6. Freeman RS, Estus S, Johnson Jr EM. Analysis of cell cycle-related gene expression in postmitotic neurons: Selective induction of Cyclin D l during programmed cell death. Neuron 1994; 12(2):343-355.

204

Hedgehog-Gli

Signaling in Human

Disease

7. Kruman II, Wersto RP, Cardozo-Pelaez F et al. Cell cycle activation linked to neuronal cell death initiated by D N A damage. Neuron 2004; 4 l ( 4 ) : 5 4 9 - 5 6 1 . 8. Chenn A, McConnell SK. Cleavage orientation and the asymmetric inheritance of Notch 1 immunoreactivity in mammalian neurogenesis. Cell 1995; 82(4):631-641. 9. Klein AL, Zilian O , Suter U et al. Murine n u m b regulates granule cell maturation in the cerebellum. Dev Biol 2004; 2 6 6 ( l ) : 1 6 l - 1 7 7 . 10. Pardee AB. G l events and regulation of cell proliferation. Science 1989; 246(4930):603-608. l l . A n g l e y C, Kumar M, Dinsio KJ et al. Signaling by bone morphogenetic proteins and Smadl modulates the postnatal differentiation of cerebellar cells. J Neurosci 2003; 23(l):260-268. 12. Li W , Cogswell CA, LoTurco JJ. Neuronal differentiation of precursors in the neocortical ventricular zone is triggered by BMP. J Neurosci 1998; 18(21):8853-8862. 13. Z h u G, Mehler M F , Zhao J et al. Sonic hedgehog and BMP2 exert opposing actions on proliferation and differentiation of embryonic neural progenitor cells. Dev Biol 1999; 215(1):118-129. 14. Solecki DJ, Liu XL, T o m o d a T et al. Activated Notch2 signaling inhibits differentiation of cerebellar granule neuron precursors by maintaining proliferation. Neuron 2 0 0 1 ; 31(4):557-568. 15. Barnabe-Heider F, Miller F D . Endogenously produced neurotrophins regulate survival and differentiation of cortical progenitors via distinct signaling pathways. J Neurosci 2003; 23(12):5149-5160. 16. Doetsch F, Petreanu L, Caille I et al. EGF converts transit-amplifying neurogenic precursors in the adult brain into multipotent stem cells. Neuron 2002; 36(6): 1021-1034. 17. Ikeya M , Lee SM, Johnson JE et al. W n t signalling required for expansion of neural crest and C N S progenitors. Nature 1997; 389(6654):966-970. 18. McMahon AP, Ingham P W , Tabin CJ. Developmental roles and clinical significance of hedgehog signaling. Curr T o p Dev Biol 2003; 53:1-114. 19. Dahmane N , Ruiz i Altaba A. Sonic hedgehog regulates the growth and patterning of the cerebellum. Development 1999; 126(14):3089-3100. 20. Wallace VA. Purkinje-cell-derived Sonic hedgehog regulates granule neuron precursor cell proliferation in the developing mouse cerebellum. Curr Biol 1999; 9(8):445-448. 2 1 . Wechsler-Reya RJ, Scott M P . Control of neuronal precursor proliferation in the cerebellum by Sonic Hedgehog [see comments]. Neuron 1999; 22(1):103-114. 22. Jensen AM, Wallace VA. Expression of Sonic hedgehog and its putative role as a precursor cell mitogen in the developing mouse retina. Development 1997; 124(2):363-371. 2 3 . Dahmane N , Sanchez P, Gitton Y et al. T h e Sonic Hedgehog-gli pathway regulates dorsal brain growth and tumorigenesis. Development 2 0 0 1 ; 128(24):5201-5212. 24. Britto J, Tannahill D , Keynes R. A critical role for sonic hedgehog signaling in the early expansion of the developing brain. Nat Neurosci 2002; 5(2):103-110. 25. Lai K, Kaspar BK, Gage FH et al. Sonic hedgehog regulates adult neural progenitor proliferation in vitro and in vivo. N a t Neurosci 2003; 6(l):21-27. 26. Machold R, Hayashi S, Rutlin M et al. Sonic hedgehog is required for progenitor cell maintenance in telencephalic stem cell niches. Neuron 2003; 39(6):937-950. 27. Schwartz MA, Baron V. Interactions between mitogenic stimuli, or, a thousand and one connections. Curr O p i n Cell Biol 1999; 11 (2): 197-202. 28. Rhoads RE. Signal transduction pathways that regulate eukaryotic protein synthesis. J Biol Chem 1999; 274(43):30337-30340. 29. Kelly K, Cochran B H , Stiles C D et al. Cell-specific regulation of the c-myc gene by lymphocyte mitogens and platelet-derived growth factor. Cell 1983; 35(3 Pt 2):603-610. 30. Lau LF, Nathans D . Identification of a set of genes expressed during the GO/Gl transition of cultured mouse cells. Embo J 1985; 4(12):3145-3151. 3 1 . Kretzschmar M , Massague J. SMADs: Mediators and regulators of TGF-beta signaling. Curr Opin Genet Dev 1998; 8(1):103-111. 32. Willert K, Nusse R. Beta-catenin: A key mediator of W n t signaling. Curr Opin Genet Dev 1998; 8(1):95-102. 3 3 . Prober DA, Edgar BA. Growth regulation by oncogenes—new insights from model organisms. Curr Opin Genet Dev 2 0 0 1 ; l l ( l ) : 1 9 - 2 6 . 34. D a n g CV. c-Myc target genes involved in cell growth, apoptosis, and metabolism. Mol Cell Biol 1999; 19(1):1-11. 35. Bouchard C, Thieke K, Maier A et al. Direct induction of cyclin D 2 by Myc contributes to cell cycle progression and sequestration of p27. Embo J 1999; 18(19):5321-5333. 36. Geng Y, Weinberg RA. Transforming growth factor beta effects on expression of G l cyclins and cyclin-dependent protein kinases. Proc Natl Acad Sci USA 1993; 90(21):10315-10319. 37. Grana X, Reddy EP. Cell cycle control in mammalian cells: Role of cyclins, cyclin dependent kinases (CDKs), growth suppressor genes and cyclin-dependent kinase inhibitors (CKIs). Oncogene 1995; 11(2):211-219.


205

38. Kenney A M , Rowitch D H . Sonic hedgehog promotes G ( l ) cydin expression and sustained cell cycle progression in mammalian neuronal precursors [In Process Citation]. Mol Cell Biol 2000; 20(23):9055-9067. 39. Oliver T G , Grasfeder LL, Carroll AL et al. Transcriptional profiling of the Sonic hedgehog response: A critical role for N - m y c in proliferation of neuronal precursors. Proc Natl Acad Sci USA 2003; 100(12):7331-7336. 40. Sewing A, Burger C, Brusselbach S et al. H u m a n cyclin D l encodes a labile nuclear protein whose synthesis is directly induced by growth factors and suppressed by cyclic A M P . J Cell Sci 1993; 104(Pt 2):545-555. 4 1 . MuUer D , Bouchard C, Rudolph B et al. Cdk2-dependent phosphorylation of p27 facilitates its Myc-induced release from cyclin E/cdk2 complexes. Oncogene 1997; 15(21):2561-2576. 42. Cheng M , Olivier P, Diehl JA et al. T h e p 2 1 ( C i p l ) and p27(Kipl) C D K 'inhibitors' are essential activators of cyclin D-dependent kinases in murine fibroblasts. Embo J 1999; 18(6):1571-1583. 43. Sherr CJ, Roberts J M . Inhibitors of mammalian G l cyclin-dependent kinases. Genes Dev 1995; 9(10):1149-1163. 44. Geng Y, Whoriskey W , Park MY et al. Rescue of cyclin D l deficiency by knockin cyclin E. Cell 1999; 97{iS)'.7G7-777. 45. Lundberg AS, Weinberg RA. Functional inactivation of the retinoblastoma protein requires sequential modification by at least two distinct cyclin-cdk complexes. M o l Cell Biol 1998; 18(2):753-761. 46. Fantl V, Stamp G, Andrews A et al. Mice lacking cyclin D l are small and show defects in eye and mammary gland development. Genes Dev 1995; 9(19):2364-2372. 47. Sicinski P, Donaher JL, Parker SB et al. Cyclin D l provides a link between development and oncogenesis in the retina and breast. Cell 1995; 82 (4):621-630. 48. Sicinski P, Donaher JL, Geng Y et al. Cyclin D 2 is an FSH-responsive gene involved in gonadal cell proUferation and oncogenesis. Nature 1996; 384(6608):470-474. 49. Sicinska E, Aifantis I, Le Cam L et al. Requirement for cyclin D 3 in lymphocyte development and T cell leukemias. Cancer Cell 2003; 4(6):451-46l. 50. Hunter T , Pines J. Cyclins and cancer. Cell 1991; 66(6):1071-1074. 5 1 . Huard J M , Forster C C , Carter M L et al. Cerebellar histogenesis is disturbed in mice lacking cyclin D 2 . Development 1999; 126(9):1927-1935. 52. Ciemerych MA, Kenney A M , Sicinska E et al. Development of mice expressing a single D-type cyclin. Genes Dev 2002; l6(24):3277-3289. 53. Ohtsubo M, Theodoras A M , Schumacher J et al. H u m a n cyclin E, a nuclear protein essential for the G l - t o - S phase transition. Mol Cell Biol 1995; 15(5):2612-2624. 54. Sherr CJ. G l phase progression: Cycling on cue. Cell 1994; 79(4):551-555. 55. Miyazawa K, Himi T, Garcia V et al. A role for p27/Kipl in the control of cerebellar granule cell precursor proliferation. J Neurosci 2000; 20(15):5756-5763. 56. O k u d a M, H o r n H F , Tarapore P et al. Nucleophosmin/B23 is a target of CDK2/cyclin E in centrosome duplication. Cell 2000; 103(1):127-140. 57. Zhao J, Kennedy BK, Lawrence BD et al. N P A T links cyclin E-Cdk2 to the regulation of replication-dependent histone gene transcription. Genes Dev 2000; l4(18):2283-2297. 58. Arata Y, Fujita M , Ohtani K et al. Cdk2-dependent and -independent pathways in E2F-mediated S phase induction. J Biol C h e m 2000; 275(9):6337-6345. 59. Geng Y, Yu Q , Sicinska E et al. Cyclin E ablation in the mouse. Cell 2003; 114(4):431-443. 60. Zarkowska T , Mittnacht S. Differential phosphorylation of the retinoblastoma protein by G l / S cychn-dependent kinases. J Biol Chem 1997; 272(19): 12738-12746. 6 1 . Sherr CJ. Cancer cell cycles. Science 1996; 274(5293):1672-1677. 62. Lee KY, Ladha M H , M c M a h o n C et al. T h e retinoblastoma protein is linked to the activation of Ras. Mol Cell Biol 1999; 1 9 ( l l ) : 7 7 2 4 - 7 7 3 2 . 63. Gupta S, Luong MX, Bleuming SA et al. T u m o r suppressor pRB functions as a corepressor of the CCAAT displacement protein (CDP/cut) to regulate cell cycle controlled histone H 4 transcription. J Cell Physiol 2003; 196(3):54l-556. 64. Valarche I, Tissier-Seta JP, Hirsch M R et al. T h e mouse homeodomain protein Phox2 regulates Ncam promoter activity in concert with C u x / C D P and is a putative determinant of neurotransmitter phenotype. Development 1993; 119(3):881-896. 65. Quaggin SE, Heuvel GB, Golden K et al. Primary structure, neural-specific expression, and chromosomal localization of Cux-2, a second murine homeobox gene related to Drosophila cut. J Biol Chem 1996; 271(37):22624-22634. GG. Ren S, RolUns BJ. Cyclin C/cdk3 promotes Rb-dependent GO exit. Cell 2004; 117(2):239-251. 67. Jiang Z , Zacksenhaus E, Gallie BL et al. T h e retinoblastoma gene family is differentially expressed during embryogenesis. Oncogene 1997; 14(15):1789-1797.

206

Hedgehog-Gli

Signaling in Human

Disease

68. Mulligan G, Jacks T. T h e retinoblastoma gene family: Cousins with overlapping interests. Trends Genet 1998; l4(6):223-229. 69. Weinberg RA. T u m o r suppressor genes. Science 1991; 254(5035):1138-1146. 70. Hamel W , Westphal M, Shepard H M . Loss in expression of the retinoblastoma gene product in human gliomas is associated with advanced disease. J Neurooncol 1993; 16(2):159-165. 7 1 . Clarke AR, Maandag ER, van Roon M et al. Requirement for a functional Rb-1 gene in murine development. Nature 1992; 359(6393):328-330. 72. Jacks T , Fazeli A, Schmitt EM et al. Effects of an Rb mutation in the mouse. Nature 1992; 359(6393):295-300. 7 3 . Lee EY, C h a n g CY, H u N et al. Mice deficient for Rb are nonviable and show defects in neurogenesis and haematopoiesis. Nature 1992; 359(6393):288-294. 74. Lee EY, H u N , Yuan SS et al. Dual roles of the retinoblastoma protein in cell cycle regulation and neuron differentiation. Genes Dev 1994; 8(17):2008-2021. 75. Macleod KF, H u Y, Jacks T . Loss of Rb activates both p53-dependent and independent cell death pathways in the developing mouse nervous system. Embo J 1996; 15(22):6178-6188. IG. Helms A W , Johnson JE. Progenitors of dorsal commissural interneurons are defined by M A T H l expression. Development 1998; 125(5):919-928. 77. Stone D M , Hynes M , Armanini M et al. T h e tumour-suppressor gene patched encodes a candidate receptor for Sonic hedgehog. Nature 1996; 384(660 5): 129-134. 78. Hallonet M E , Teillet MA, Le Douarin N M . A new approach to the development of the cerebellum provided by the quail-chick marker system. Development 1990; 108(1):19-31. 79. Goodrich LV, Scott M P . Hedgehog and patched in neural development and disease. Neuron 1998; 21(6):1243-1257. 80. Kalyani AJ, Piper D , Mujtaba T et al. Spinal cord neuronal precursors generate multiple neuronal phenotypes in culture. J Neurosci 1998; 18(19):7856-7868. 8 1 . Rowitch D H , B SJ, Lee SM et al. Sonic hedgehog regulates proliferation and inhibits differentiation of C N S precursor cells. J Neurosci 1999; 19(20):8954-8965. 82. H o KS, Scott M P . Sonic hedgehog in the nervous system: Functions, modifications and mechanisms. Curr O p i n Neurobiol 2002; 12(l):57-63. 83. Marigo V, Davey RA, Zuo Y et al. Biochemical evidence that patched is the Hedgehog receptor. Nature 1996; 384(6605):176-179. 84. Ingham PW. Transducing Hedgehog: T h e story so far. Embo J 1998; 17(13):3505-3511. 85. Epstein DJ, Marti E, Scott M P et al. Antagonizing cAMP-dependent protein kinase A in the dorsal C N S activates a conserved Sonic h e d g e h o g signaling p a t h w a y . D e v e l o p m e n t 1996; 122(9):2885-2894. 86. Fan C M , Porter JA, Chiang C et al. Long-range sclerotome induction by sonic hedgehog: Direct role of the amino-terminal cleavage product and modulation by the cyclic A M P signaling pathway. Cell 1995; 81(3):457-465. 87. Hammerschmidt M , Bitgood MJ, McMahon AP. Protein kinase a is a common negative regulator of Hedgehog signaling in the vertebrate embryo. Genes Dev 1996; 10(6):647-658. 88. Hammerschmidt M, McMahon AP. T h e effect of pertussis toxin on zebrafish development: A possible role for inhibitory G-proteins in hedgehog signaling. Dev Biol 1998; 194(2):166-171. 89. Ungar AR, Moon RT. Inhibition of protein kinase a phenocopies ectopic expression of hedgehog in the C N S of wild-type and cyclops mutant embryos. Dev Biol 1996; 178(1):186-191. 90. D e C a m p DL, Thompson T M , de Sauvage FJ et al. Smoothened activates Galphai-mediated signaling in frog melanophores. J Biol Chem 2000; 275(34):26322-26327. 9 1 . Murone M , Rosenthal A, de Sauvage FJ. Sonic hedgehog signaling by the patched-smoothened receptor complex. Curr Biol 1999; 9(2):76-84. 92. Nybakken KE, T u r c k C W , Robbins DJ et al. Hedgehog-stimulated phosphorylation of the kinesin-related protein Costal2 is mediated by the serine/threonine kinase fused. J Biol Chem 2002; 277(27):24638-24647. 9 3 . Monnier V, H o KS, Sanial M et al. Hedgehog signal transduction proteins: Contacts of the fused kinase and Ci transcription factor with the kinesin-related protein Costal2. B M C Dev Biol 2002; 2(1):4. 94. Sisson J C , H o KS, Suyama K et al. Costal2, a novel kinesin-related protein in the Hedgehog signaling pathway. Cell 1997; 90(2):235-245. 95. Robbins DJ, Nybakken KE, Kobayashi R et al. Hedgehog elicits signal transduction by means of a large complex containing the kinesin-related protein costal2. Cell 1997; 90(2):225-234. 96. Delia Rocca GJ, Maudsley S, Daaka Y et al. Pleiotropic coupling of G protein-coupled receptors to the mitogen-activated protein kinase cascade. Role of focal adhesions and receptor tyrosine kinases. J Biol Chem 1999; 274(20): 13978-13984.


207

97. Luttrell LM, Daaka Y, Lefkowitz RJ. Regulation of tyrosine kinase cascades by G-protein-coupled receptors. Curr O p i n Cell Biol 1999; 11 (2): 177-183. 98. Post GR, Brown J H . G protein-coupled receptors and signaling pathways regulating growth responses. Faseb J 1996; 10(7):74l-749. 99. Gutkind JS. T h e pathways connecting G protein-coupled receptors to the nucleus through divergent mitogen-activated protein kinase cascades. J Biol C h e m 1998; 273(4):1839-1842. 100. Goodrich LV, Milenkovic L, Higgins KM et al. Altered neural cell fates and medulloblastoma in mouse patched mutants. Science 1997; 277(5329):1109-1113. 101. Ingham PW, M c M a h o n AP. Hedgehog signaling in animal development: Paradigms and principles. Genes Dev 2 0 0 1 ; 15(23):3059-3087. 102. H a h n H , Wicking C, Zaphiropoulous PG et al. Mutations of the human homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome. Cell 1996; 85(6):841-851. 103. RaflFel C, Jenkins RB, Frederick L et al. Sporadic meduUoblastomas contain P T C H mutations. Cancer Res 1997; 57(5):842-845. 104. Wetmore C, Eberhart D E , Curran T . Loss of p53 but not ARF accelerates medulloblastoma in mice heterozygous for patched. Cancer Res 2 0 0 1 ; 61(2):513-516. 105. Marino S, Vooijs M , van Der Gulden H et al. Induction of meduUoblastomas in p53-null mutant mice by somatic inactivation of Rb in the external granular layer cells of the cerebellum. Genes Dev 2000; 14(8):994-1004. 106. Pomeroy SL, Tamayo P, Gaasenbeek M et al. Prediction of central nervous system embryonal tumour outcome based on gene expression. Nature 2002; 415(6870):436-442. 107. Rubin JB, Rowitch D H . Medulloblastoma: A problem of developmental biology. Cancer Cell 2002; 2(l):7-8. 108. Banker GaG K. Culturing Nerve Cells. Cambridge MA: M I T Press, 1998. 109. Durand B, RaiFM. A cell-intrinsic timer that operates during oligodendrocyte development. Bioessays 2000; 22(1):64-71. 110. Ruiz i Altaba A. GH proteins and Hedgehog signaling. Trends in Genetics 1999; 15(10):4l8-425. 111. Sasaki H , Nishizaki Y, H u i C et al. Regulation of Gli2 and Gli3 activities by an amino-terminal repression domain: Implication of Gli2 and Gli3 as primary mediators of Shh signaling. Development 1999; 126(17):3915-3924. 112. Dai P, Akimaru H , Tanaka Y et al. Sonic Hedgehog-induced activation of the glil promoter is mediated by GLI3. J Biol Chem 1999; 274(12):8143-8152. 113. Bai C B , Joyner AL. G l i l can rescue the in vivo function of G l i 2 . D e v e l o p m e n t 2 0 0 1 ; 128(24):5161-5172. 114. W a n g B, Fallon JF, Beachy PA. Hedgehog-regulated processing of Gli3 produces an anterior/posterior repressor gradient in the developing vertebrate Umb. Cell 2000; 100(4):423-434. 115. Kinzler K W , Bigner SH, Bigner D D et al. Identification of an amplified, highly expressed gene in a human gUoma. Science 1987; 236(4797):70-73. 116. Ruiz i Altaba A, Sanchez P, Dahmane N . Gli and hedgehog in cancer: T u m o u r s , embryos and stem cells. N a t Rev Cancer 2002; 2(5):361-372. 117. Yoon J W , Kita Y, Frank DJ et al. Gene expression profiling leads to identification of GLI 1-binding elements in target genes and a role for multiple downstream pathways in GLI 1-induced cell transformation. J Biol C h e m 2002; 277(7):5548-5555. 118. Park HL, Bai C, Piatt KA et al. Mouse Glil mutants are viable but have defects in S H H signaling in combination with a Gli2 mutation. Development 2000; 127(8):1593-1605. 119. Weiner H L , Bakst R, Hurlbert MS et al. Induction of meduUoblastomas in mice by sonic hedgehog, independent of GUI. Cancer Res 2002; 62(22):6385-6389. 120. Bai CB, Stephen D , Joyner AL. All mouse ventral spinal cord patterning by hedgehog is Gli dependent and involves an activator function of Gli3. Dev Cell 2004; 6(1): 103-115. 121. Henriksson M , Luscher B. Proteins of the Myc network: Essential regulators of cell growth and differentiation. Adv Cancer Res 1996; 68:109-182. 122. Mateyak MK, Obaya AJ, Adachi S et al. Phenotypes of c-Myc-deficient rat fibroblasts isolated by targeted homologous recombination. CeU Growth Differ 1997; 8(10):1039-1048. 123. Garson JA, P e m b e r t o n LF, Sheppard P W et al. N - m y c gene expression a n d o n c o p r o t e i n characterisation in medulloblastoma. Br J Cancer 1989; 59(6):889-894. 124. Stanton BR, Perkins AS, TessaroUo L et al. Loss of N-myc function results in embryonic lethality a n d failure of t h e epithelial c o m p o n e n t of t h e e m b r y o to d e v e l o p . G e n e s D e v 1 9 9 2 ; 6(12A):2235-2247. 125. Sawai S, Shimono A, Hanaoka K et al. Embryonic lethality resulting from disruption of both N-myc alleles in mouse zygotes. N e w Biol 1991; 3(9):861-869.

208

Hedgehog-Gli

Signaling in Human

Disease

126. Charron J, Malynn BA, Fisher P et al. Embryonic lethality in mice homozygous for a targeted disruption of the N-myc gene. Genes Dev 1992; 6(12A):2248-2257. 127. Moens CB, Auerbach AB, Conlon RA et al. A targeted mutation reveals a role for N-myc in branching morphogenesis in the embryonic mouse lung. Genes Dev 1992; 6(5):691-704. 128. Knoepfler PS, Cheng PF, Eisenman RN. N-myc is essential during neurogenesis for the rapid expansion of progenitor cell populations and the inhibition of neuronal differentiation. Genes Dev 2002; 16(20):2699-2712. 129. Kenney AM, Widlund HR, Rowitch D H . Hedgehog and PI-3 kinase signaling converge on N m y c l to p r o m o t e cell cycle progression in cerebellar n e u r o n a l precursors. D e v e l o p m e n t 2 0 0 4 ; 131(l):217-228. 130. Ayer DE, Eisenman RN. A switch from Myc: Max to Mad: Max heterocomplexes accompanies monocyte/macrophage differentiation. Genes Dev 1993; 7(11):2110-2119. 131. HurUn PJ, Queva C, Koskinen PJ et al. M a d 3 and Mad4: Novel Max-interacting transcriptional repressors that suppress c-myc dependent transformation and are expressed during neural and epidermal differentiation. Embo J 1995; l4(22):5646-5659. 132. Schreiber-Agus N , Chin L, Chen K et al. An amino-terminal domain of Mxil mediates anti-Myc oncogenic activity and interacts with a homolog of the yeast transcriptional repressor SIN3. Cell 1995; 80(5):777-786. 133. W u M, Arsura M , Bellas RE et al. Inhibition of c-myc expression induces apoptosis of W E H I 231 murine B cells. Mol Cell Biol 1996; l6(9):5015-5025. 134. Lahoz EG, Xu L, Schreiber-Agus N et al. Suppression of Myc, but not E l a, transformation activity by Max- associated proteins, Mad and M x i l . Proc Natl Acad Sci USA 1994; 91(12):5503-5507. 135. Kenney A M , Cole M D , Rowitch D H . Nmyc upregulation by sonic hedgehog signaling promotes proliferation in developing cerebellar granule neuron precursors. Development 2003; 130(l):15-28. 136. Bazarov AV, Adachi S, Li SF et al. A modest reduction in c-myc expression has minimal effects on cell growth and apoptosis but dramatically reduces susceptibility to Ras and Raf transformation. Cancer Res 2 0 0 1 ; 61(3):1178-1186. 137. Prendergast G C , Lawe D , Ziff EB. Association of Myn, the murine homolog of max, with c-Myc stimulates methylation-scnsitive D N A binding and ras cotransformation. Cell 1991; 65(3):395-407. 138. McMahon SB, Van Buskirk HA, Dugan KA et al. T h e novel ATM-related protein TRRAP is an essential cofactor for the c- Myc and E2F oncoproteins. Cell 1998; 94(3):363-374. 139. Park J, Kunjibettu S, McMahon SB et al. T h e ATM-related domain of TRRAP is required for histone acetyltransferase r e c r u i t m e n t and M y c - d e p e n d e n t oncogenesis. Genes Dev 2 0 0 1 ; 15(13):1619-1624. 140. Lasorella A, Boldrini R, Dominici C et al. Id2 is critical for cellular proliferation and is the oncogenic effector of N-myc in human neuroblastoma. Cancer Res 2002; 62(l):301-306. 141. Galaktionov K, Chen X, Beach D . Cdc25 cell-cycle phosphatase as a target of c-myc. Nature 1996; 382(6591):511-517. 142. Claassen GF, H a n n SR. Myc-mediated transformation: The repression connection. Oncogene 1999; 18(19):2925-2933. 143. Malynn BA, de Alboran IM, O'Hagan R C et al. N-myc can functionally replace c-myc in murine development, cellular growth, and differentiation. Genes Dev 2000; 14(11): 1390-1399. 144. Albert T , Urlbauer B, Kohlhuber F et al. Ongoing mutations in the N-terminal domain of c-Myc affect transactivation in Burkitt's lymphoma cell lines. Oncogene 1994; 9(3):759-763. 145. Henriksson M , Bakardjiev A, Klein G et al. Phosphorylation sites mapping in the N-terminal domain of c-myc modulate its transforming potential. Oncogene 1993; 8(12):3199-3209. 146. Sears R, Nuckolls F, Haura E et al. Multiple Ras-dependent phosphorylation pathways regulate Myc protein stability. Genes Dev 2000; 14(19):2501-2514. 147. Niklinski J, Claassen G, Meyers C et al. Disruption of Myc-tubulin interaction by hyperphosphorylation of c-Myc during mitosis or by constitutive hyperphosphorylation of mutant c-Myc in Burkitt's lymphoma. Mol Cell Biol 2000; 20(l4):5276-5284. 148. Salghetti SE, Kim SY, Tansey W P . Destruction of Myc by ubiquitin-mediated proteolysis: Cancer-associated and transforming mutations stabilize Myc. Embo J 1999; 18(3):717-726. 149. Cohen P, Frame S. T h e renaissance of GSK3. Nat Rev Mol Cell Biol 2 0 0 1 ; 2(10):769-776. 150. Jia J, Amanai K, W a n g G et al. Shaggy/GSK3 antagonizes Hedgehog signalling by regulating Cubitus interruptus. Nature 2002; 4l6(6880):548-552. 151. Lutterbach B, H a n n SR. c-Myc transactivation domain-associated kinases: Questionable role for map kinases in c-Myc phosphorylation. J Cell Biochem 1999; 72(4):483-491. 152. Dudek H , Datta SR, Franke T F et al. Regulation of neuronal survival by the serine-threonine protein kinase Akt. Science 1997; 275(5300):661-665.


209

153. Ye P, Xing Y, Dai Z et al. In vivo actions of insulin-like growth factor-I (IGF-I) on cerebellum development in transgenic mice: Evidence that IGF-I increases proliferation of granule cell progenitors. Brain Res Dev Brain Res 1996; 95(l):44-54. 154. de Pablo F, de la Rosa EJ. T h e developing C N S : A scenario for the action of proinsulin, insulin and insulin-like growth factors. Trends Neurosci 1995; 18(3): 143-150. 155. Schmidt M , Fernandez de Mattos S, van der Horst A et al. Cell cycle inhibition by FoxO forkhead transcription factors involves downregulation of cyclin D . Mol Cell Biol 2002; 22(22):7842-7852. 156. Klein RS, Rubin JB, Gibson H D et al. SDF-1 alpha induces chemotaxis and enhances Sonic hedgehog-induced proliferation of cerebellar granule cells. Development 2 0 0 1 ; 128(11):1971-1981. 157. Kho AT, Zhao Q , Cai Z et al. Conserved mechanisms across development and tumorigenesis revealed by a mouse development perspective of human cancers. Genes Dev 2004; 18(6):629-640. 158. Wechsler-Reya R, Scott M P . T h e developmental biology of brain tumors. A n n u Rev Neurosci 2 0 0 1 ; 24:385-428. 159. Frank-Kamenetsky M , Zhang X M , Bottega S et al. Small-molecule modulators of Hedgehog signaling: Identification and characterization of Smoothened agonists and antagonists. J Biol 2002; 1(2):10. 160. Taipale J, Chen JK, Cooper M K et al. Effects of oncogenic mutations in Smoothened and Patched can be reversed by cyclopamine [see comments]. Nature 2000; 406(6799): 1005-1009. 161. Stecca B, Ruiz i Altaba A. T h e therapeutic potential of modulators of the Hedgehog-Gli signaling pathway. J Biol 2002; 1(2):9. 162. Berman D M , Karhadkar SS, Hallahan AR et al. MeduUoblastoma growth inhibition by hedgehog pathway blockade. Science 2002; 297(5586):1559-1561. 163. Wetmore C. Sonic hedgehog in normal and neoplastic proliferation: Insight gained from human tumors and animal models. Curr O p i n Genet Dev 2003; 13(l):34-42. 164. Roy S, Ingham PW. Hedgehogs tryst with the cell cycle. J Cell Sci 2002; 115(Pt 23):4393-4397. 165. Del Valle L, Enam S, Lassak A et al. Insulin-like growth factor I receptor activity in human meduUoblastomas. Clin Cancer Res 2002; 8(6):1822-1830. 166. Sano T, Lin H , Chen X et al. Differential expression of M M A C / P T E N in glioblastoma multiforme: Relationship to localization and prognosis. Cancer Res 1999; 59(8): 1820-1824. 167. Yeh E, Cunningham M , Arnold H et al. A signalling pathway controlling c-Myc degradation that impacts oncogenic transformation of human cells. N a t Cell Biol 2004; 6(4):308-318. 168. Zhao Q , Kho A, Kenney A M et al. Identification of genes expressed with temporal-spatial restriction to developing cerebellar neuron precursors by a functional genomic approach. Proc Natl Acad Sci USA 2002; 99(8):5704-5709. 169. Sim AT, Collins E, Mudge LM et al. Developmental regulation of protein phosphatase types 1 and 2A in post-hatch chicken brain. Neurochem Res 1998; 23(4):487-491. 170. Santa-Coloma TA. Anp32e ( C p d l ) and related protein phosphatase 2 inhibitors. Cerebellum 2003; 2(4):310-320. 171. Lu KP, Liou YC, Zhou X Z . Pinning down proline-directed phosphorylation signaling. Trends Cell Biol 2002; 12(4):164-172. 172. Ryo A, Liou YC, Lu KP et al. Prolyl isomerase P i n l : A catalyst for oncogenesis and a potential therapeutic target in cancer. J Cell Sci 2003; 116(Pt 5):773-783. 173. Lu PJ, Z h o u X Z , Liou YC et al. Critical role of W W domain phosphorylation in regulating phosphoserine binding activity and P i n l function. J Biol C h e m 2002; 277(4):2381-2384. 174. Derynck R, Zhang YE. Smad-dependent and Smad-independent pathways in TGF-beta family signalling. Nature 2003; 425(6958):577-584. 175. Seoane J, Le H V , Massague J. Myc suppression of the p 2 1 ( C i p l ) C d k inhibitor influences the outcome of the p53 response to D N A damage. Nature 2002; 4l9(6908):729-734. 176. Waite KA, Eng C. B M P 2 exposure results in decreased P T E N protein degradation and increased P T E N levels. H u m Mol Genet 2003; 12(6):679-684. 177. Sakamuro D , Elliott KJ, Wechsler-Reya R et al. B I N l is a novel MYC-interacting protein with features of a tumour suppressor. N a t Genet 1996; l 4 ( l ) : 6 9 - 7 7 . 178. Nikiforov MA, Chandriani S, Park J et al. TRRAP-dependent and TRRAP-independent transcriptional activation by Myc family oncoproteins. Mol Cell Biol 2002; 2 2 ( l 4 ) : 5 0 5 4 - 5 0 6 3 .

CHAPTER 16

Modulating the Hedgehog Pathway in Diseases Frederic J. de Sauvage* and Lee L. Rubin Introduction

T

he hedgehog (Hh) pathway is a signaling system that regulates a wide range of develop mental processes.^ Hh proteins act as morphogens to induce cell differentiation in a dose dependent manner, control cell proliferation, or alter cell shape. Aberrant regulation of the Hh signal, caused by mutation in components of the pathway, has been shown to lead to various developmental disorders^ or more recendy to cancer.^' Most components of the Hh pathway were first identified in flies and later shown to be conserved in higher organisms. Although still incomplete, a picture of how the hedgehog signal is transmitted at the cellular level has started to emerge, allowing for potential therapeutic approaches to the treatment of diseases involving this pathway. Hedgehog proteins undergo autoproteolytic cleavage to generate an NH2-terminal fragment that is cholesterol modified at its COOH-terminus and palmitoylated at its NH2-terminus.^ Secretion and diffusion of this dually lipid modified protein requires the activity of Dispatched,^ a multi-transmembrane protein with homology to Patched (see below), and heparan sulfate proteoglycans,^ a major component of extracellular matrix. Reception of the signal transmitted by Hh or its three mammalian orthologs, sonic hedgehog (Shh), desert hedgehog (Dhh) and Indian hedgehog (Ihh), is mediated by two multitransmembrane cell surface proteins. Patched (Ptch) and Smoothened (Smo). Ptch is a twelve transmembrane domain protein with features similar to those of a transporter, which, in the absence of ligand, represses the activity of Smo, a seven-transmembrane protein. Upon binding of Hh to Ptch, the inhibition of Smo is relieved and the Hh signal is transmitted via a protein complex that includes a number of components thought to be specific to the pathway: Costal2 (cos2), an atypical kinesin-like protein. Fused, a serine-threonine protein kinase, Suppressor of Fused (SuFu) and the zinc-fingers transcription factor cubitus interruptus (Ci) or Glil ,2 and 3 in mammals. Additional proteins, including PKA, GSK3p and C K l a also play a critical role in regulating the activity and nuclear localization of Glis (for review see re£ 8). Germline mutations in the PTCHl gene lead to a condition known as basal cell nevus syndrome (BCNS or Gorlin syndrome) where patients develop large numbers of basal cell carcinomas (BCC) on their skin and suffer higher incidence of other tumor types such as meduUoblastoma.^'^^ Most sporadic BCC tumors have high Hh pathway activity, often as a result of mutation in P7U//followed by loss of heterozygosity. Approximately 10% of sporadic BCCs do not have a PTCHl mutation, but harbor a specific heterozygous mutation in SMOy ^^ which makes SMO less sensitive to P T C H l mediated inhibition. Rare potential gain of *Corresponding Author: Frederic J. de Sauvage—Department of Molecular Biology, Genentech, Inc. South San Francisco, California, U.S.A. Email: [email protected].

Hedgehog-GU Signaling in Human Disease, edited by Ariel Ruiz i Altaba. ©2006 Eurekah.com and Springer Science+Business Media.

Modulating the Hedgehog Pathway in Diseases

Hh-Ag

211

Cyclopamine

Cur~61414 H3C PH.,

CH3/ H \

0

i IN ^^

/

< Jl J

NMi

Figure 1. Small molecules modulator of the Hh pathway. function mutations in 57///itself have been described but it has vet to be demonstrated how they could contribute to increased Hh activity and tumorigenesis. Mutations in downstream signaling components of the pathway may also contribute to increased Hh signaling activity and tumor formation. However, to date only a few instances have been documented. GLIl was initially discovered as an amplified oncogene in human glioma but this appears to be a rare event. More recently, mutations in SUFU, a negative regulator of the pathway involved in tethering Glis in the cytoplasm, were identified in some cases of familial medulloblastomas^^ suggesting that SUFU may function as a tumor suppressor. In addition to some tumors bearing mutations in SHH or downstream Hh pathway components, other tumor cells may themselves synthesize high levels of Hh ligand and rely on its autocrine or paracrine activities for growth or survival. Small cell lung carcinomas, as well as tumors derived from the upper gastrointestinal tract, particularly pancreatic tumors were shown to require increased Hh signaling mosdy through ligand stimulation. Indeed, these tumors were shown in various in vitro and in vivo experiments to be sensitive to inhibition of Hh ligand by a neutralizing monoclonal antibody or by cyclopamine (Fig. 1), a steroidal alkaloid which binds to SMO and specifically inhibits the Hh pathway. Together, these observations indicate that agents capable of inhibiting the Hh pathway at the level of Hh itself, or preferably downstream of PTCH, at the level of SMO could have an application in the treatment of Hh-dependent cancers.

Identification of Small Molecule Hedgehog (Hh) Antagonists The observation that steroidal alkaloids such as cyclopamine and jervine inhibit Hh signaling suggested that it should be possible to identify additional small molecules that modulate the Hh pathway. To identify such small molecules, approximately 200,000 compounds were screened in a cell-based assay for activation (see below) or inhibition of Hh activity. In this assay, a stable mouse lOTl/2 cell line expressing luciferase under the control of a Gli-1 DNA binding sequence was employed. Hh signaling was initiated by the addition of recombinant Hh protein, test compounds were added at the same time and luciferase activity was quantified after an overnight incubation. Numerous hits were identified, but since the goal of the screen was to identify potential therapeutics, and not simply useful rccênts to probe the Hh pathway, they were prioritized by their drug-like nature. Even with this added filter, several quite potent chemical classes of antagonists were identified. The first objective of the Hh antagonist program was to derive a topical treatment for BCC. One compound class, represented by the aminoproline Cur-61414 (Fig. 1), was highly soluble and seemed likely to have sufficient permeability to penetrate well into the skin lesions. This antagonist does not inhibit frizzled,^ the Wnt receptor that shows the greatest sequence similarity to smoothened. In other studies, Cur-6l4l4 was found not to compete in assays that

212


measure the binding of numerous other Ugands to their cognate G-protein coupled receptor. Therefore, Cur-6l4l4 appears to be a specific antagonist of the Hh pathway. Cur-6l4l4 has been tested in a variety of standard in vitro activities to confirm its abiUty to inhibit Hh signaling but, most cogendy, has been tested in two cell culture models of BCC. In the first model, Williams et al recapitidated a previously described transgenic mouse model of BCC where Shh was overexpressed in basal keratinocytes ^ by exposing embryonic mouse skin punches to recombinant Shh. In this model, Cur-6l4l4 was able to prevent the formation of BCC-like lesions produced after several days of constitutive activation of Hh signaling. Cur-61414 also cause the disappearance of preformed lesions via decreased proliferation and increased apoptosis.^ Importandy, both effects were restricted to cells in which the Hh pathway was hyper-activated, leaving surrounding normal skin unaffected. Similar effects were observed using another BCC model. Aszterbaum et al showed that long-term irradiation of Ptch heterozygous mice (a model of Gorlin syndrome) resulted in BCC-like skin lesions. These mice were UV-irradiated for more than 6 months and lesion-bearing skin punches were placed in culture. Again treatment of skin explants from the mice with Cur-6l4l4 produced a selective induction of apoptosis in the lesions without obvious effects on adjacent skin. Using the screening procedure described above, an additional class of potent small molecule Hh antagonist intended to be used systemically, was identified.^^ In this case, it was important that these antagonists be sufficiendy potent and preferably orally available. The compounds were also optimized for plasma half-life and other properties associated with effective tumor therapy and have now been shown to be effective in inhibiting growth and inducing apoptosis of meduUoblastoma cells in vitro and in vivo.

Identification of Hh Small Molecule Hh Agonists The best-characterized function of the Hh pathway is to establish the number and type of cells in the brain and spinal cord during embryonic development.^^'^^ For example, spinal cord motor neurons and midbrain dopaminergic neurons both owe their existence to the activity of Shh. A small number of experiments have confirmed that this pathway continues to have the potential to contribute to the regulation of cell number and cell-type in the adult nervous system. For example, Lai et al have shown that over-expression of Shh in the adult rat brain continues to stimulate stem cell proliferation and neural differentiation.^^ Other studies concluded that, when injected into the striatum, Shh protein could enhance the function of dopaminergic neurons in models of Parkinson's disease. These data raised the possibility that activators of the Hh pathway might be of therapeutic value in treating disorders of the nervous system by promoting cell survival, maintaining neural differentiation or by accelerating the appearance of new neurons after neuronal damage. Since viable central nervous system (CNS) therapeutics are most likely to be small molecules, a reporter gene screens similar to the one used to identify small molecule Hh antagonists was employed to identify potential agonists of the Hh pathway. One of the weak hits identified in the screen was subjected to extensive medicinal chemistry, culminating in the synthesis of sub-nM EC50 compounds with good oral availability and CNS penetration. Interestingly, even weak agonists coidd stimulate proliferation of cerebellar granule neural progenitors and induce differentiation of ventral neural cell-types in spinal cord explants, both known Hh-mediated activities.^^ Additionally, it was demonstrated that retinoic acid-pretreated mouse ES cells could be stimulated to differentiate into motor neurons with either recombinant Shh or with the small molecule Hh agonists.^^ Continued chemical modification of the Hh-Ag class led to the synthesis of compounds that were more active, on a molar basis, than the normal Hh protein ligands. Understanding the Hh-Ag mechanism of action therefore became quite important. Genetic epistasis studies suggested that Hh-Ag targeted a component of the Hh signaling pathway at or near the level of Smo. Direct cross-linking and, eventually, binding experiments using radiolabeled Hh-Ag

Modulating the Hedgehog Pathway in Diseases

213

confirmed that it bound to Smo, confirming the hypothesis that Smo is a readily draggable component of the Hh signaling pathway. Once orally available agonists able to readily cross the blood-brain barrier were identified, they were tested in various CNS disease models. Surprisingly, they have proven to be effective in decreasing the amount of neuronal cell death in different excitotoxicity models, including stroke. Given 6 hours afi:er middle cerebral artery occlusion (a standard stroke model), the Hh agonists are able to markedly decrease the size of the infarct that result from the ischemia and subsequent reperfiision. The mechanism of action underlying this neuroprotective effect is not entirely clear at this point. When given orally, even in a single dose, the agonists stimulate proliferation of neural stem cells in the hippocampus and subventricular zone.^"^ That is, the agonists offer the potential ability to stimulate the appearance of new neurons in the adult nervous system. In the context of a stroke-damaged brain, this could be of significant therapeutic consequence. Thus, small molecule Hh agonists offer the potential of being both neuroprotective and neuroregenerative.

Conclusion Recent progress in the understanding of the Hedgehog pathway and the development of molecules that modulate its activity opens new avenues for the treatment of a number of devastating diseases. Inhibitors of the pathway, small molecules or antibodies may prove useful for highly aggressive cancers such as pancreatic tumors or meduUoblastoma, the most common childhood brain tumor. On the other hand, small molecule agonists of the pathway provide new therapeutic options for the treatment of neurodegenerative disorders such as Parkinson s disease.

References 1. Ingham PW, M c M a h o n AP. Hedgehog signaling in animal development: Paradigms and principles. Genes & Development 2 0 0 1 ; 15:3059-3087. 2. McMahon AP, Ingham P W , Tabin CJ. Developmental roles and clinical significance of hedgehog signaling. Curr T o p Dev Biol 2003; 53:1-114. 3. Pasca di Magliano M , Hebrok M . Hedgehog signalling in cancer formation and maintenance. Nat Rev Cancer 2003; 3:903-11. 4. Taipale J, Beachy P. T h e Hedgehog and W n t signalling pathways in cancer. Nature 2001:349-354. 5. M a n n RK, Beachy PA. Novel lipid modifications of secreted protein signals. Annu Rev Biochem 2004; 73:891-923. 6. Burke R, Nellen D , Bellotto M et al. Dispatched, a novel sterol-sensing domain protein dedicated to the release of cholesterol-modified hedgehog from signaling cells. Cell 1999; 99:803-15. 7. Bellaiche Y, T h e I, Perrimon N . Tout-velu is a Drosophila homologue of the putative tumour suppressor EXT-1 and is needed for H h diffusion. Nature 1998; 394:85-8. 8. Lum L, Beachy PA. T h e Hedgehog response network: Sensors, switches, and routers. Science 2004; 304:1755-9. 9. H a h n H , Wicking C, Zaphiropoulous P C et al. Mutations of the human homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome. Cell 1996; 85:841-51. 10. Johnson RL, Rothman AL, Xie J et al. H u m a n homolog of patched, a candidate gene for the basal cell nevus syndrome. Science 1996; 272:1668-71. 11. Unden AB, Kerstin B, Zaphiropulos P C et al. T h e gene for Gorlin's syndrome, human patched, is consistently overexpressed in both hereditary and sporadic basal cell carcinomas. Journal of investigative dermatology 1997; 108:596. 12. Xie J, Murone M , Luoh SM et al. Activating smoothened mutations in sporadic basal-cell carcinoma. Nature 1998; 391:90-2. 13. Couve-Privat S, Le Bret M , Traiffort E et al. Functional analysis of novel sonic hedgehog gene mutations identified in basal cell carcinomas from xeroderma pigmentosum patients. Cancer Res 2004; 64:3559-65. 14. Kinzler KW, Bigner SH, Bigner D D et al. Identification of an amplified, highly expressed gene in a human glioma. Science 1987; 236:70-3. 15. Taylor M D , Liu L, Raffel C et al. Mutations in SUFU predispose to meduUoblastoma. Nature Genetics 2002; 31:306-10.

214


16. Berman DM, Karhadkar SS, Maitra A et al. Widespread requirement for Hedgehog iigand stimulation in growth of digestive tract tumours. Nature 2003; 425:846-51. 17. Thayer SP, di MagUano MP, Reiser PW et al. Hedgehog is an early and late mediator of pancreatic cancer tumorigenesis. Nature 2003; 425:851-6. 18. Watkins DN, Berman DM, Burkholder SG et al. Hedgehog signalling within airway epithelial progenitors and in small-cell lung cancer. Nature 2003; 422:313-7. 19. Chen JK, Taipale J, Cooper MK et al. Inhibition of Hedgehog signaling by direct binding of cyclopamine to smoothened. Genes Dev 2002; 16:2743-8. 20. Frank-Kamenetsky M, Zhang XM, Bottega S et al. Small-molecule modulators of Hedgehog signaling: Identification and characterization of smoothened agonists and antagonists. J Biol 2002; 1:10. 21. Oro AE, Higgins KM, Hu Z et al. Basal cell carcinomas in mice overexpressing sonic Hedgehog. Science 1997; 276:817-21. 22. Williams JA, Guicherit OM, Zaharian BI et al. Identification of a small molecule inhibitor of the hedgehog signaling pathway: Effects on basal cell carcinoma-like lesions. Proc Natl Acad Sci USA 2003; 100:4616-21. 23. Aszterbaum M, Epstein J, Oro A et al. Ultraviolet and ionizing radiation enhance the growth of BCCs and trichoblastomas in patched heterozygous knockout mice. Nat Med 1999; 5:1285-91. 24. Romer JT, Kimura H, Magdaleno S et al. Suppression of the Shh pathway using a small molecule inhibitor eUminates medulloblastoma in Ptcl+/- p53-/- mice. Cancer Cell 2004; 6:229-240. 25. Jessell TM. Neuronal specification in the spinal cord: Inductive signals and transcriptional codes. Nat Rev Genet 2000; 1:20-9. 26. Rallu M, Machold R, Gaiano N et al. Dorsoventral patterning is established in the telencephalon of mutants lacking both Gli3 and Hedgehog signaling. Development 2002; 129:4963-74. 27. Ye W, Shimamura K, Rubenstein J et al. FGF and Shh signals control dopaminergic and serotonergic cell fate in the anterior neural plate. Cell 1998; 93:755-766. 28. Lai K, Kaspar BK, Gage FH et al. Sonic hedgehog regulates adult neural progenitor proliferation in vitro and in vivo. Nat Neurosci 2003; 6:21-7. 29. Tsuboi K, Shults CW. Intrastriatal injection of sonic hedgehog reduces behavioral impairment in a rat model of Parkinson's disease. Exp Neurol 2002; 173:95-104. 30. Wichterle H, Lieberam I, Porter JA et al. Directed difiverentiation of embryonic stem cells into motor neurons. Cell 2002; 110:385-97. 31. Dellovade TL, Bless E, Albers DS et al. Efficacy of small-molecule hedgehog agonists in models of excitotoxicity. Soc Neurosci 2003; 33:446-7. 32. Machold R, Hayashi S, Rutlin M et al. Sonic hedgehog is required for progenitor cell maintenance in telencephalic stem cell niches. Neuron 2003; 39:937-50.

CHAPTER 17

Hedgehog Signaling in Endodermally Derived Tumors Marina Pasca di Magliano and Matthias Hebrok* The Hedgehog Signaling Pathway

A

relatively small number of intercellular signaling pathways, including the Hedgehog (Hh), transforming growth factor P (TGF-p), fibroblast growth factor (FGF), Wnt, and Notch pathways, interact to regulate embryonic development and organogenesis. In contrast to the other pathways, the mammalian Hedgehog signaling pathway is relatively compact, comprised of only three known ligands. Sonic (Shh), Indian (Ihh), and Desert Hedgehog (Dhh), and two receptors, patched 1 and 2 (Ptchl/2; (Fig. l)?^ In the absence of the ligands. Patched (Ptch) inhibits the activity of a second transmembrane protein, Smoothened (Smo). As a consequence, transcription factors of the Gli family remain inactive in the cytoplasm through interaction with a protein complex that includes Suppressor of fused su(fu). Upon ligand stimulation, Ptch mediated inhibition of Smo is alleviated and a cascade of events that is not fully understood results in the activation and translocation of Gli transcription factors into the nucleus where they activate the transcription of target genes. Some of these are components of the Hedgehog pathway itself, including Gli, Ptch and the Hedgehog interacting protein (Hhip). Both Ptch and Hhip are membrane proteins that attenuate the pathway by binding to and blocking diffusion of Hedgehog ligands. '^^'^^' Therefore, the Hedgehog signaling pathway regulates itself through a negative feedback mechanism that results in increased expression of Ptch and Hhip in Hedgehog responsive cells.

Hedgehog Signaling in Adult Endodermal Organs The importance of a functional Hedgehog signaling pathway during embryogenesis becomes apparent in transgenic mice lacking Smo^ a component of the pathway through which all signaling appears to be mediated. Only one Smo gene has been identified in vertebrates and homozygous mutant embryos die shordy after gastrulation. Similarly, loss or reduction of Hedgehog activity has been implicated with developmental defects in a variety of diflFerent organs (see ref. 23). Here, we will focus on the role of the Hedgehog pathway during the formation of endodermal organs. While all three Hedgehog ligands are expressed in endodermal cells during embryogenesis, no apparent defects in endodermal tissues have been noted in mice carrying homozygous mutations in Dhh. By contrast, mice that lack either Sonic or Indian Hedgehog display numerous defects along the gastrointestinal tract^^' ^ as well as loss of the branching morphogenesis in the lung.^^'^ Increasing evidence suggests that Hedgehog signaling functions in a concentration-dependent manner and different levels of the pathway

*Corresponding Author: Matthias Hebrok—Diabetes Center, Department of Medicine, University of California San Francisco, 513 Parnassus Ave., San Francisco, California 94143 U.S.A. Email: [email protected]


216


Figure 1. Hedgehog (Hh) signaling pathway. In the absence ofHgands, Patched (Ptch) suppresses the activity of Smoothened (Smo), another transmembrane protein that is essential for Hedgehog signal transduction. As a consequence, Gli transcription factors, the downstream mediators of the Hedgehog pathway, are kept inactive within a cytoplamic complex that includes suppressor of fused (Su(fu)). Upon ligand binding, Ptch mediated inhibition of Smo is alleviated and Smo activation results in signals that lead to Gli translocation into the nucleus. Here, Gli factors activate target gene transcription, including Ptchy Hhip and Gli itself. Both Ptch and Hhip have been shown to attenuate Hedgehog signaling via sequestering of Hedgehog ligands (Hh), including Sonic, Indian, and Desert Hedgehog. have been implicated in generating molecular boundaries required for proper organogenesis. In the foremidgut area, both Shh and Ihh are expressed in forming stomach and duodenal tissues while the endodermal cells developing into pancreatic structures are devoid of both ligands at early stages. ' Ectopic activation of Hedgehog signaling via tissue manipulation and forced expression of Shh under control of the Pancreatic and duodenal homeobox gene 1 {Pdx-1) promoter results in loss of pancreatic marker genes and transformation of pancreatic mesenchyme into duodenal mesoderm.^' Interestingly, stepwise increase in Hedgehog activity in transgenic mice carrying mutations in Hhip or Hhip and Ptch alleles, both inhibitors of Hedgehog signaling, result in increasingly severe phenotypes in lung and pancreas,^'"^ indicating that Hedgehog pathway activity has to be tightly controlled to allow proper organ formation. While a lot of progress has been made in understanding which embryonic tissues respond to Hedgehog signaling, much less is known about the role of hedgehog signaling in adult tissues. A number of recent studies have shown that the pathway remains active in a subset of adult organs, mostly confined to a subset of cells within each organ. At least in some organs, the activity of the pathway in mature tissues appears to be lower than what has been shown in embryonic tissues (Fig. 2), thus posing a potential problem in identifying cells that receive and respond to Hedgehog ligands. However, previous reports had shown that Ptch, the receptor and antagonist of the pathway, is also a transcriptional target and thus a faithful and quantitative marker of Hedgehog-responsive cells. Most studies have used transgenic mice in which the bacterial LacZ gene has been used to replace the Ptch gene.^^ While the homozygous Ptch knock-out is early embryonic lethal, the heterozygous mice survive and staining for galactosidase activity allows, identification of cells with active Hedgehog signaling. Interestingly, the Hedgehog pathway is down-regulated in endodermal organs and only few cells are marked by

Hedgehog Signaling in Endodermally Derived Tumors

217

Lung B.

h^ "€Sw^%i

Pancreas E.

D.

k \ .

*.

^^'.'^^Wk HjflWi^^"y\

% \

.Mm

Colon

G. i'.r^5.

•

1

.

:

Figure 2. Hedgehog signaling in adult endodermal organs. A-C) Lung. A) Shh expression is undetectable in normal lung tissue. B) Upon tissue damage induced by naphtalene treatment a subset of cells becomes Shh positive. C) In contrast to normal tissue, small cell lung cancers (SCLC) are marked by significant expression of Shh. D-F) Pancreas. D) Similar to lung, expression of Shh is undetectable by immunohistochemistry in normal human pancreas. E) Pancreatic intraepithelial lesions are marked by a noticeable increase in Shh expression levels and the majority of pancreatic adenocarcinoma show high levels of Shh expression (F). G-I) Colon. G,H) By contrast to lung and pancreas, expression of Ihh is maintained in adult colon where it is confined to differentiated colonocytes. I) Colonic cancer have lost Ihh expression, suggesting that Hedgehog signaling in colon tissue might have anti-oncogenic fiinctions. The arrow marks Ihh-positive, differentiated colonocytes in G,H, and colonocytes in colon adenoma that have down-regulated Ihh expression in I. Figures 2 a-c were originally published in Watkins et al. Nature 2003; 422:313-317; Figures 2 d-f were originally published in Thayer at al. Nature 2003; 425:851-856; Figures 2 g-i were originally published in van den Brink at al. Nature Genetics 2004; 36:277-282. All the figures were reproduced with permission from the Nature Publishing Group. p-galactosidase activity. In the adult lung, Puh is expressed in small clusters of cells, believed to constitute the neuroendocrine precursors within the airway epithelium, in the basal layer of the bronchial epithelium. ^^ Neither Shh nor Gli are detectable in the adult lung. In the stomach and gall bladder some Ptch positive cells are found in the epithelium. Ptch expression is observed at verv low levels in the murine pancreatic ducts and the endocrine cells of the islet of Langerhans. ' However, immunohistochemistry is not sensitive enough to detect Ptch, Shh, and Smo in h u m a n pancreatic tissues. An exciting possibility is that Hedgehog signaling marks a population of progenitor cells in several endodermal organs. While this hypothesis awaits confirmation from cell isolation and transplantation studies, publications analyzing other adult organs, mainly the brain, have noted

218


that Hedgehog signaling promotes cell proliferation of progenitor cells to maintain tissue homoeostasis. Intriguingly, tissue damage seems to activate Hedgehog signaling in these cells, suggesting that the level of pathway activation directs the proliferative response in these cells. In the brain, Shh controls proliferation of cerebellum granule neuron precursors^ ^'^^ as well as the proliferation of progenitors of the tectum and neocortex. ^^ Moreover, Hh signaling in the postnatal telencephalon is required to promote proliferation and maintain the populations of progenitor cells. '^^

Hedgehog Signaling in Cancers ofEndodermal Origin Lung The normal role of Hedgehog signaling in subsets of cells in mature tissues is a matter of intense research and could lead to important insights into how embryonic signaling pathways are reemployed to maintain tissue homeostasis. Growing evidence suggests that deregulated Hedgehog signaling within adult tissues can initiate or support tumor formation in mice and humans. In general, increased pathway activity, either through sporadic mutation (Smo, Su(fu)) or in familial conditions (Gorlins syndrome, mutation in Ptch), has been shown to lead to tumor formation in the skin, the cerebellum and skeletal muscle. ' ' More recent publications have addressed the role of Hedgehog signaling in tumors of endodermally derived organs, such as lungs and organs in the gastrointestinal tract. Lung cancer, the leading cause of cancer death world wide,"^ is classified as small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC) based on the histological appearance of the cancer cells. Small cell lung cancer is a highly lethal malignancy believed to derive from the Hedgehog-positive neuroendocrine cells, a specialized progenitor population located within the airway epithelium (Fig. 2). In normal airway epithelium, small clusters of Ptch positive cells are observed that do not stain for Shh or Gli expression. Following lung damage, a process during which specialized cells of the lung epithelium (Clara cells) are depleted, the expression of both Shh and Gli is up regulated and putative airway progenitor cells start to proliferate. In SCLC, Hedgehog signaling molecules can be detected, both in primary human tumors and in tumor-derived cell lines. Furthermore, SCLC cells secrete Shh ligand and inhibition of ligand-receptor binding with anti-Shh blocking antibodies inhibits pathway activation, suggesting that Hedgehog signaling regulates its activity via an autocrine loop in these cells. In addition, SCLCs are susceptible to treatment with cyclopamine, an alkaloid derived from a plant of the lily family (Veratrum californicum) that physically binds to Smoothened and inhibits Hedgehog signal transduction at this level. Cyclopamine-mediated inhibition results in reduction of cell survival and prevents tumor growth. Thus, the observation that cyclopamine treatment blocks the growth of the cancer cells opens up the possibility to develop new therapeutic approaches that rely on inhibition of the Hh pathway. However, it is important to note that Hedgehog signaling is usually not active in non-small cell lung cancer (NSCLC). Thus, it is likely that only distinct cell types within the adult lung, and most likely other organs, carry the ability to respond to upregulation of Hedghog ligand expression with increased proliferation while other cells types remain quiescent. This also suggests that the growing number of Hedgehog-antagonists will only block tumor progression in a subset of cancer types (Fig. 3).

Hedgehog Signaling in Gastrointestinal Tumors of the Pancreas and Colon Recently, activation of the hedgehog pathway has been associated with several tumors of gastrointestinal origin. ' Tumors originating from oesophagus, stomach, biliary tract, and pancreas show increased Hedgehog activity. Most tumor cell lines derived from tumors of these tissues express both Shh and Ihh Hedgehog ligands. Similar to the situation found in SCLCs, cell proliferation can be inhibited using either anti-Shh/anti-Ihh blocking antibodies or cyclopamine while purified ligands stimulate cell replication. It is noteworthy that in contrast to the other gastrointestinal tumors, colon tumors are not marked by increased Hedgehog


219

anti-Shh blocking antibody

cyclopamine KAAD-cyclop, Cur61414

Ptch, Gli ^ antisense

Figure 3. Hedgehog signaling inhibitors. An increasing number of Hedgehog inhibitors has been identified that block Hedgehog activation at different levels.^^ Inhibition ofHedgehog signal transduction could result in novel therapeutic treatments for endoderm-derived tumors. signaling. Below, we will discuss the different mechanisms by which Hedgehog signaling might influence cancer formation and growth by comparing its role in pancreatic adenocarcinoma and colon tumors.

Pancreas Pancreatic adenocarcinoma is a devastating disease for which no effective therapy is currently available.^ The origin of the progenitor cell that forms this cancer is still controversial, however, the epithelial nature and expression of duct markers suggests that tumors might derive from pancreatic duct cells. Histological analysis of a large number of tumor patients has resulted in a progression model of precancerous lesions, known as pancreatic intraepithelial neoplasia (PanIN 1-3).'^^ During progression from normal duct epithelium to carcinoma in situ, pancreatic duct cells first loose their normal cuboidal morphology and cell polarity, accumulate mucin, and finally evaginate from the epithelium to metastasize into the adjacent organs. The different PanIN lesions are also characterized by the accumulation of genetic mutations in tumor suppressor genes, including K-ras, DPC4, and pi6.^ Immunohistochemical analysis has shown that Hedgehog signaling components are overexpressed in a significant percentê of PanIN 1 lesions, suggesting that deregulation of the pathway is one of the early steps during pancreatic cancer formation. ^ Additional evidence for increased Hedgehog signaling as an initiating event comes from studies of transgenic mice that express Shh under control of the Pdx-1 promoter. Pancreas formation in these mice is severely compromised and transgenic animals usually die within the first weeks after birth. Nonetheless, forced Hedgehog activation results in morphological changes reminiscent of human PanIN lesions. Analysis of a large series of pancreatic cancer cells lines derived either from primary or metastatic adenocarcinomas shows that all lines express components of the Hedgehog signaling pathway, indicating that the pathway is still active at later stages of tumor formation. Inhibition of the

220


pathway with cyclopamine blocks proUferation and induces apoptosis in a subset of the cancer ceil lines, demonstrating an essential role for Hedgehog signaling during tumor growth and survival. While cyclopamine is extremely effective in blocking proliferation of a subset of pancreatic cancer cell lines, about half of the cell lines do not respond to this treatment. It is currendy unclear whether the cyclopamine resistant lines do not depend on Hedgehog activity for their growth or whether mutations within the pathway downstream of Smoothened render them insensitive to this drug. In different tumors, such as medulloblastoma and basal cell carcinoma, Ptch is often mutated or deleted, leading to constitutive activation of Smoothened.'^'^ Moreover, gain-of-fimction mutations of Smoothened have been identified.^ In some cases, mutations in suppressor of fused, another negative regulator of the pathway, have been found in medulloblastoma. ^ So far, gastrointestinal cancer, including pancreatic adenocarcinoma, have not been analyzed for mutations in Hedgehog signaling components and it is thus unclear whether such mutations are implicated in familial cases of these tumors. It is interesting to note that all GI derived tumors and cell lines analyzed were found to express, and depend on the continuous presence of, Hedgehog ligands. The possibility that these tumors ensure their growth and survival via an autocrine signaling loop might provide clues for therapeutic interventions.

Colon As a general rule, ectopic activation rather than silencing of Hedgehog signaling has been implicated in tumor formation.^ ' ^ One possible exception is colon cancer in which Hedgehog signaling might play a protective role by counteracting the activity of another embryonic signaling pathway, the Wnt pathway. In normal colon tissue, progenitor cells are located at the base of a crypt structure. These cells proliferate to self renew and to give rise to cells that undergo terminal differentiation, a process that depends on the presence of Indian Hedgehog. Thus, in contrast to other gastrointestinal tissues. Hedgehog signaling does not induce cell proliferation but cell differentiation, van den Brink and colleagues have recendy shown that Ihh functions at least in part by repressing the activity of the Wnt pathway known to be deregulated in colon cancer cells. Similarly, Wnt signaling represses Ihh expression, thereby establishing a delicate balance between progenitor cells higher in Wnt activity and differentiated cells higher in Hedgehog activity (Fig. 4). The importance of this delicate balance is confirmed by the opposite phenotypes of two knock-out mice: in Tcf4'' mice, which lack active Wnt signaling in the colon, the progenitor cell compartment gets rapidly depleted after birth, as all the cells in the colon initiate terminal differentiation. By contrast, Ihh knock-out embryos are characterized by actively proliferating cells that fail to differentiate into mature colonocytes.^^ While this scenario provides an elegant explanation of how mutual inhibiton of the Wnt and Hedgehog pathways allows simultaneous maintenance of undifferentiated, progenitor cells and differentiated colonocytes other studies have noted expression of both Ihh and Shh transcripts at the base of the developing crypts during embryogenesis. Future studies will have to address whether adult crypt cells produce active Ihh and Shh protein and whether Hedgehog signaling provides different clues for developing versus adult colon tissue. Emerging evidence suggests that the balance between Wnt and Hedgehog signaling is also implicated in formation of colon cancer. Colon cancer cell lines are negative for Hedgehog signaling activity and Ihh expression is lost in colon polyps and tumors. ' Numerous studies have shown that increase in Wnt signaling, marked by nuclear p-catenin, is sufficient to cause colon cancer. As expected from the mutual inhibition of Hedgehog and Wnt signaling observed in normal colon tissue, the cells that have lost Ihh expression show staining of beta-catenin within the nucleus. Thus, it is possible that Ihh reduction, normally required to ensure proper differentiation of colonocytes through inhibition of overt Wnt signaling, would allow uncontrolled proliferation of crypt progenitor cells. According to this hypothesis. Hedgehog signaling in colon cancer would work as a tumor suppressor rather than an oncogenic pathway (Fig. 4). While intriguing, further research is needed to address some of the unresolved issues. For


221

A. Wnt-Hh interactions in colon tissue and colon cancer: differentiated

Ihh

colonocyte

(ihh) differentiation

/

differentiation

•

Wnt Progenitor cell

IP^ ^ [jij ^ ^

^ ^ Wnt

C^ J ^ f r b - f ^ ^ [ i f i ^ J ^ ^

uncontrolled proliferation

Self renewal proliferationi Norma I Colon

Colon Cancer

B. Other endodermal organs: 1

Wnt?

Hh

1 ^ ^

uncontrolled proliferation

proliferation Norm a I tiss ue

Ca nee r

Figure 4. Hedgehog/Wnt interactions in gastrointestinal organs. A) Recent studies suggest that Wnt and Hedgehog signaling antagonize each other in colon to maintain homeostasis of undifferentiated progenitor cells (Wnt positive) and differentiated cell types (Hedgehog positive). A change in this balance, e.g., through activating mutations in the Wnt pathway, results in reduction ofIhh expression and subsequently an increase in progenitor cell proliferation that can lead to colon cancer. B) By contrast, increased Hedgehog signaling has been implicated in tumor formation in other endodermal organs, including esophagus, lung, stomach, and pancreas. Thus, the normal role of Hedgehog signaling in these organs might be to maintain proliferative capacities of a subset of cells. The role ofWnt signaling in these organs and the interaction between the Hedgehog and Wnt pathway have not been elucidated. example, it would be interesting to determine whether colon cancer cells carry mutations in Hedgehog signaling components that reduce the activity of the pathway. Also, a subset of colon cancer cell lines do express Ptch and Gliy transcriptional targets of Hedgehog signaling and it will be important to analyze whether W n t and Hedgehog signaling behave in antagonistic fashion in all colon cancers. Furthermore, it might be important to understand whether cell proliferation and differentiation in other organs, and therefore ultimately t u m o r formation, is also regidated via this inhibitory feedback between the W n t and Hedgehog pathways (Fig. 4).

222


Hedgehog Interactions with Other Pathways During embryogenesis Hedgehog and Wnt signaling have been shown to activate each others activity in various organs and recent evidence suggests that similar interactions also occur in adult tissues and tumors. ' ' An example for positive interactions between these pathways comes from studies of cerebellum-derived medulloblastoma. MeduUoblastoma are often caused by misregulation of the Hh pathway and are also marked by active Wnt signaling. Similarly, nuclear beta-catenin has been detected in several cases of basal cell carcinomas, another tumor known to be caused by ectopic activation of the Hedgehog pathway. Additional evidence comes from the observation that several Wnt pathway components are upregulated when BCC-like skin tumors are induced in late xenopus tadpoles by expression of human GUI ?^ Thus, the reinforcing and antagonizing regulation of the Wnt and Hedgehog in different tumors suggest that the nature of these interactions is tissue dependent. Previous studies have noted the similarities between the Wnt and Hedgehog pathways.^^ In both pathways, the ligands are lipid modified and Smoothened and Frizzleds, the Wnt receptors, are related proteins. Furthermore, intracellular signaling components, e.g., glycogen synthase kinase 3 (GSK-3), appear to be shared by both pathways. While the exact mechanism that regulates the interaction between the two pathways is as yet unknown, one intriguing possibility involves the fiinction of one component of the Hedgehog pathway. Suppressor of fiised (Su(fu)), an inhibitory component of the Hh signaling pathway, is known to bind Gli transcription factors and prevent their activation. ^ ^ ' ' More recendy, Su(fti) has also been shown to bind beta-catenin, and prevent it from entering the nucleus.^^ Translocation of beta-catenin to the nucleus is essential for Wnt activation, as it leads to the formation of a p-catenin-TCF transcription factor complex that activates transcription of the Wnt target genes. Thus, Su(fij) might act as a negative regulator of both the Wnt and the Hh pathway. Interestingly, a mutation of Su(fu),which is not able to bind beta-catenin, has been described in medulloblastoma,"^^' thereby providing a possible link between the activation of both pathways in this type of cancer. Evidence for a positive interaction between both pathways in endodermal organ-derived tumors is currendy missing and ftiture studies might address this question. Both the Notch and the Hh pathway play an important role in pancreatic cancer. ' ' While direct interactions between these pathways in pancreatic tumors have not been reported, circumstantial evidence suggests a link between both pathways. Within pancreatic tissue, EGF receptor signaling has been shown to activate the Notch pathway; in a positive feedback loop, the Notch pathway is required to activate TGFa, a ligand for the EGF receptor.^"^ Studies addressing the relation between Hh and EGF signaling in mammalian tumors are currendy missing, however, recent evidence indicates that Shh and EGF cooperate to regulate the proliferation of neural stem cells.^ In addition, Hh signaling is known to induce expression of EGF receptors and ligands in Drosophila. Thus, it is tempting to speculate that Hedgehog signaling might regulate Notch activity in pancreatic and other gastrointestinal cancers via activation of the EGF signaling pathway. This would be particularly intriguing as Notch induction has been associated with the activation of P-catenin and Lef-1, markers of active Wnt signaling, in skin. Understanding the interactions between embryonic signaling pathways is important as more and more pathway inhibitors become available. In the best case scenario, distinct combinations of pathway inhibitors will provide novel therapeutic options for treatment of diverse cancers.

Acknowledgements We would like to thank the members of the Hebrok group for stimulating discussions and critical reading of the chapter. We would like to thank Dr G. van den Brink, Dr N. Watkins and the Nature Publishing group for permission to reprint figures from published articles. Work in M.H. s laboratory is supported by grants from the NIH, Juvenile Diabetes Research Foundation (JDRF), Hillblom Foundation, and Lustgarten Foundation.

Hedgehog Signaling in Endodermally

Derived

Tumors

223

References 1. Amin A, sLi Y, Finkelstein R. Hedgehog activates the EGF receptor pathway during Drosophila head development. Development 1999; 126:2623-30. 2. Apelqvist A, Ahlgren U, Edlund H . Sonic hedgehog directs specialised mesoderm differentiation in the intestine and pancreas. Curr Biol 1997; 7:801-804. 3. Bardeesy N , DePinho RA. Pancreatic cancer biology and genetics. Nat Rev Cancer 2002; 2:897-909. 4. Berman D M , Karhadkar SS, Maitra A et al. Widespread requirement for Hedgehog ligand stimulation in growth of digestive tract tumours. Nature 2003; 425:846-851. 5. Bitgood MJ, McMahon AP. Hedgehog and BMP genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev Biol 1995; 172:126-138. 6. Bitgood MJ, Shen L, McMahon AP. Sertoli cell signaling by Desert hedgehog regulates the male germline. Curr Biol 1996; 6:298-304. 7. Chen JK, Taipale J, Cooper M K et al. Inhibition of Hedgehog signaling by direct binding of cyclopamine to Smoothened. Genes and Development 2002; 16:2743-8. 8. Chen Y, Struhl G. Dual roles for patched in sequestering and transducing Hedgehog. Cell 1996; 87:553-63. 9. Chuang PT, Kawcak T , McMahon AP. Feedback control of mammalian Hedgehog signaling by the Hedgehog-binding protein, H i p l , modulates Fgf signaling during branching morphogenesis of the lung. Genes and Development 2003; 17:342-7. 10. Chuang P T , M c M a h o n AP. Vertebrate Hedgehog signalling modulated by induction of a Hedgehog- binding protein. Nature 1999; 397:617-21. 11. Dahmane N , Ruiz i Altaba A. Sonic hedgehog regulates the growth and patterning of the cerebellum. Development 1999; 126:3089-100. 12. Dahmane N , Sanchez P, Gitton Y et al. T h e sonic Hedgehog-gli pathway regulates dorsal brain growth and tumorigenesis. Development 2 0 0 1 ; 128:5201-5212. 13. Ding Q, Fukami S, Meng X et al. Mouse suppressor of fused is a negative regulator of sonic hedgehog signaling and alters the subcellular distribution of G l i l . Curr Biol 1999; 9:1119-22. 14. Doble BW, Woodgett JR. GSK-3: Tricks of the trade for a multi-tasking kinase. J Cell Sci 2003; 116:1175-86. 15. Goodrich LV, Johnson RL, Milenkovic L et al. Conservation of the hedghog/patched signaling pathway from flies to mice: Induction of a mouse patched gene by hedgehog. Genes & Dev 1996; 10:301-312. 16. Goodrich LV, Jung D , Higgins KM et al. Overexpression of p t c l inhibits induction of Shh target genes and prevents normal patterning in the neural tube. Developmental Biology 1999; 211:323-34. 17. Goodrich LV, Milenkovic L, H i ^ i n s KM et al. Altered neural cell fates and meduUoblastoma in mouse patched mutants. Science 1997; 277:1109-13. 18. Hebrok M . Hedgehog signaling in pancreas development. Mech Dev 2003; 20:45-57. 19. Hebrok M , Kim SK, Melton DA. Notochord repression of endodermal sonic hedgehog permits pancreas development. Genes and Development 1998; 12:1705-1713. 20. Hebrok M , Kim St SK, Jacques B et al. Regulation of pancreas development by Hedgehog signaling. Development 2000; 127:4905-4913. 2 1 . Hruban R H , Adsay N V , Albores-Saavedra J et al. Pancreatic intraepithelial neoplasia: A new nomenclature and classification system for pancreatic duct lesions. Am J Surg Pathol 2 0 0 1 ; 25:579-86. 22. Ingham P W . T h e patched gene in development and cancer. Curr O p i n Genet Dev 1998; 8:88-94. 23. Ingham P W , M c M a h o n A!^. Hedgehog signaling in animal development: Paradigms and principles. Genes and Development 2 0 0 1 ; 15:3059-87. 24. Kawahira H , M a N H , Tzanakakis ES et al. Combined activities of Hedgehog signaling inhibitors regulate pancreas development. Development 2 0 0 3 , in press. 25. Kerr KM. Pulmonary preinvasive neoplasia. Journal of clinical pathology 2 0 0 1 ; 54:257-271. 26. K o g e r m a n P, G r i m m T , K o g e r m a n L et al. M a m m a l i a n suppressor-of-fused m o d u l a t e s nuclear-cytoplasmic shuttUng of Gli-1. N a t Cell Biol 1999; 1:312-9. 27. Korinek V, Barker N , Moerer P et al. Depletion of epithelial stem-cell compartments in the small intestine of mice lacking Tcf-4. N a t Genet 1998; 19:379-83. 28. Lai K, Kaspar BK, Gage F H et al. Sonic hedgehog regulates adult neural progenitor proliferation in vitro and in vivo. N a t Neurosci 2003; 6:21-7. 29. Litingtung Y, Lei L, Westphal H et al. Sonic hedgehog is essential to foregut development. N a t Genet 1998; 20:58-61. 30. Machold R, Hayashi S, Rutlin M et al. Sonic hedgehog is required for progenitor cell maintenance in telencephalic stem cell niches. Neuron 2003; 39:937-50. 3 1 . Meng X, Poon R, Zhang X et al. Suppressor of fused negatively regulates beta-catenin signaling. J Biol Chem 2 0 0 1 ; 276:40113-9.

224


32. Miyamoto Y, Maitra A, Ghosh B et al. Notch mediates TGF alpha-induced changes in epithelial differentiation during pancreatic tumorigenesis. Cancer Cell 2003; 3:565-76. 33. MuUor JL, Dahmane N, Sun T et al. Wnt signals are targets and mediators of Gli function. Curr Biol 2001; 11:769-73. 34. Murone M, Luoh SM, Stone D et al. Gli regulation by the opposing activities of fused and suppressor of fused. Nat Cell Biol 2000; 2:310-2. 35. Nusse R. Wnts and Hedgehogs: Lipid-modified proteins and similarities in signaling mechanisms at the cell surface. Development 2003; 130:5297-305. 36. Palma V, Ruiz i Altaba A. Hedgehog-GLI signaling regulates the behavior of cells with stem cell properties in the developing neocortex. Development 2004; 131:337-45. 37. Parr BA, McMahon AP. Dorsalizing signal Wnt-7a required for normal polarity of D-V and A-P axes of mouse limb. Nature 1995; 374:350-3. 38. Pasca di Magliano M, Hebrok M. Hedgehog signaling in cancer formation and maintenance. Nature Reviews Cancer 2003; 3:903-911. 39. Pepicelli CV, Lewis PM, McMahon AP. Sonic hedgehog regulates branching morphogenesis in the mammalian lung. Curr Biol 1998; 8:1083-6. 40. Pinson KI, Brennan J, Monkley S et al. An LDL-receptor-related protein mediates Wnt signalling in mice. Nature 2000; 407:535-8. 41. Qualtrough D, Buda A, Gaffield W et al. Hedgehog signalling in colorectal tumour cells: Induction of apoptosis with cyclopamine treatment. Int J Cancer 2004; 110:831-837. 42. Ramalho-Santos M, Melton DA, McMahon AP. Hedgehog signals regulate multiple aspects of gastrointestinal development. Development 2000; 127:2763-2772. 43. Ruiz i Altaba A, Sanchez P, Dahmane N. Gli and hedgehog in cancer: Tumours, embryos and stem cells. Nat Rev Cancer 2002; 2:361-372. 44. Stone DM, Murone M, Luoh S et al. Characterization of the human suppressor of fused, a negative regulator of the zinc-finger transcription factor GU. J Cell Sci 1999; 112(Pt 23):4437-4448. 45. Tabata T, Kornberg TB. Hedgehog is a signaling protein with a key role in patterning Drosophila imaginal discs. Cell 1994; 76:89-102. 46. Taipale J, Chen JK, Cooper MK et al. Effects of oncogenic mutations in smoothened and patched can be reversed by cyclopamine. Nature 2000; 406:1005-1009. 47. Taylor MD, Liu L, Raffel C et al. Mutations in SUFU predispose to meduUoblastoma. Nat Genet 2002; 31:306-310. 48. Taylor MD, Zhang X, Liu L et al. Failure of a medulloblastoma-derived mutant of SUFU to suppress WNT signaling. Oncogene 2004; 23:4577-4583. 49. Thayer SP, Pasca di Magliano MP, Heiser PW et al. Hedgehog is an early and late mediator of pancreatic cancer tumorigenesis. Nature 2003; 425:851-856. 50. Thomas MK, Rastalsky N, Lee JH et al. Hedgehog signaling regulation of insulin production by pancreatic beta- cells. Diabetes 2000; 49:2039-47. 51. van den Brink GR, Bleuming SA, Hardwick JC et al. Indian Hedgehog is an antagonist of Wnt signaling in colonic epithelial cell differentiation. Nat Genet 2004; 36:277-282. 52. Wallace VA. Purkinje-cell-derived Sonic hedgehog regulates granule neuron precursor cell proliferation in the developing mouse cerebellum. Curr Biol 1999; 9:445-448. 53. Watkins DN, Berman DM, Burkholder SG et al. Hedgehog signalUng within airway epithelial progenitors and in small-cell lung cancer. Nature 2003; 422:313-317. 54. Xie J, Murone M, Luoh SM et al. Activating smoothened mutations in sporadic basal-cell carcinoma. Nature 1998; 391:90-92. 55. Zhang XM, Ramalho-Santos M, McMahon AP. Smoothened mutants reveal redundant roles for Shh and Ihh signaling including regulation of L/R symmetry by the mouse node. [Republished From: Cell. 2001 Jun 15;105(6):781-92 UI: 21334154]. Cell 2001; 106:781-792.

Index Accutane 157 Acrocapitofemoral dysplasia (ACFD) 146, 148-151 Agonist 13,28,212,213 AUele 30, 35, 53, 55-58, 60, 64-66, 70, 71, 131, 133, 139, 141, 142, 147, 195, 197, 216 Antagonist 5, 13, 27, 49, 110, 126, 178, 184, 198,203,211,212,216,218 Autosomal dominant inheritance 133

B Basal cell carcinoma (BCC) 2-6, 8, 9, 12, 13, 29, 30, 42, 46, 47, 49, 53-56, 58, 59, 63-71, 74, 75, 77-82, 92, 96, 100-107, 110,156,210-212,220,222 Basalioma 53 BCL2 76, 79, 81, 105 Bone morphogenetic protein (BMP) 41, 47, 49, 95, 100, 124, 125, 128, 146, 182, 188, 203 Brachydactyly 146, 147, 149, 150 Brain tumor 8, 58, 81, 92, 128, 177, 187, 192, 202, 203, 213

Caenorhabditis elegans 25, 187 Calcitonin gene related peptide (CGRP) 124 Cancer 1-5, 8-14, 41, 53-55, 58, 64, 65, 68, 69, 74, 80-83, 86, 101, 104, 110, 123, 125, 126, 128, 161, 191, 192, 199, 202, 203,210,211,213,217-222 CD4+T lymphocyte 119,123 CD8+T lymphocyte 123 Celebrex® 68 Cell cycle 1, 26, 46, 59, 69, 77, 79-81, 97, 99, 101, 105, 107-110, 122, 123, 126, 127, 146, 187-193, 195-203 Cell signaling 25 Central nervous system (CNS) 3, 4, 8, 9, 88, 90-93, 127, 131, 133, 157, 164, 177, 187, 188, 191-193, 195-197, 199-203, 212,213 Cerebellum 3, 30, 54, 57, 65, 88, 127, 158, 189,192-195, 197,218,222

Chick 3, 47, 49, 93, 97, 138, 142, 154, 155, 160, 161, 163-168, 178, 183 Cholesterol 6, 23, 25, 34-36, 38, 120, 157, 161,193,210 Clara cell secretory protein (CCSP)-expressing epithelial cells 124 Clinoril® 68 Colon 5, 68, 92, 120, 217-221 Colon cancer 68, 220, 221 Congenital malformation 147 Cortex (CTX) 6, 179-182, 188, 193 Craniofacial development 153, 157-159, 161, 164, 171 Cyclin 25, 59, 66, 81, 99, 107, 109, 110, 189, 191, 192, 196, 197, 199, 200, 202, 203 Cyclopamine 2, 5-13, 71, 101, 110, 120, 127, 155, 160, 161, 168-170, 184, 199, 202, 211,218,220 Cyclopia 6, 91, 153, 155, 158, 161, 162 Cyclosporine A 70

D D-type 81, 107, 109, 110, 189, 191, 196, 197, 202 Degradation 36, 39, 48, 86, 107, 108, 161, 200-202 Desert hedgehog (Dhh) 35, 87, 119, 127, 146,178,210,215,216 Development 1, 4-6, 9, 13, 14, 25, 28, 41, 46, 47, 50, 55, 58, 63-66, 68-70, 74-76, 78-83, 86, 90-102, 104, 105, 107, 108, 123, 126-128, 142, 146, 147, 150, 151, 153, 155,157-162, 164, 166-171, 177, 179-184, 187, 188, 191-202, 212, 213, 215 DifRision 26, 34-36, 38, 39, 210, 215 Digit identity 137,138 Dispatched (Disp) 23, 35, 38, 120, 210 Dorsal-ventral patterning 100, 164, 177, 178, 180-184 Drosophik 1, 3, 14, 23-27, 30, 35, 81-83, 86, 87, 98, 107, 131, 156, 177, 178, 183, 194,195,222


226

Endoderm 41, 104, 128, 161, 162, 165, 219 Epidermal stem cells 78, 96, 97, 101, 102 Epidermis 25, 26, 30, 43, 45-47, 49, 50, 64, 68, 74-82, 86, 92, 96-100, 102, 104, 105, 106, 108, 109, 164 Epidielial-mesenchymal transition (EMT) 12, 79,81 Epithelium 12, 26, 34, 35, 39, 41-49, 92, 98, 101, 103, 119-121, 123-125, 128, 162, 164, 166, 168-170, 217-219 Export 27, 35, 38

Fibroblast growth factor (FGF) 13, 41, 49, 83, 95, 121-123, 125, 127, 146, 162, 168, 178,189,215 Fibroblast growth factor 8 (FGF8) 159, 162-164, 168, 170, 178-183 Fibroblast growth factor 10 (FGFIO) 121-123, 127, 125, 168, 169 Fibrosis 119-123,126,127 FITC mouse model 121,122 Forebrain 3, 7, 30, 88, 89, 92, 98, 107, 153, 155, 158, 162, 164, 166, 168, 178, 180, 189,203

G-protein coupled receptors (GPCRs) 190, 193, 195,212 GABAergic neuron 181 Ganglionic eminence 179-183 Genotype 3, 58, 64, 65, 129, 134, 156 Genotype-phenotype correlation 3, 129, 134, 156 GLI/Gli 1-9, 11-14, 24, 48, 64, 69, 74-76, 78-83, 86-88, 92-95, 98-103, 105, 107, 108-110, 120, 124, 126-128, 131, 141, 170, 178, 190, 193, 195-197, 200, 211, 215-218,221,222 GLI3/Gli3 3, 14, 75, IG. 79-83, 87, 89-91, 93-95, 98-101, 107-109, 129-134, 137, 140-143, 156, 159, 160, 166, 170, 177, 178, 180-184, 194, 196 Glioma 3-5, 196,202,211 Glutamatergic neuron 181 Gorlin syndrome (GS) 2, 3, 29, 30, 74, 156, 195,210,212,218 Gradient 24, 26-28, 30, 34, 36, 38, 95, 100, 127, 138, 141, 162, 183, 184

Greig cephalopolysyndactyly syndrome (GCPS) 89, 129-133, 142, 143, 156, 159, 182 Growth arrest specific gene 1 (Gasl) 2, 30, 100, 162, 169

H Hair cycle 43, 48, 49, 68, 97, 98, 102, 107-109 Hair follicle 12, 13, 41-47, 49, 50, 66-68, 74-80, 86, 88, 91, 92, 95-102, 104-109 Hedgehog (HH/Hh) 1-9, 11-14, 23-28, 30, 34, 35, 36, 41, 42, 46, 53, 63-71, 74, 75, 78-83, 86, 87, 91, 94-98, 109-110, 119, 120, 122-128, 131, 137, 146, 148, 153, 157-160, 165, 168, 177, 178, 183, 187, 189-191, 193, 195-197,199-203, 210-213,215-222 Hedgehog lipid modifications 25 Hedgehog (Hh) signaling 1-6, 9, 12, 23, 24, 26, 28, 30, 34, 41, 46, 63-69, 71, 74, 83, 86,87,94,95,98,101,103,105, 107-110, 119, 120, 123, 125-128, 153, 159, 177, 178, 187, 189, 193,195-197, 200-202,211-213,215-222 Hedgehog-interacting protein (Hhip) 5, 215, 216 Heparan sulfate proteoglycan (HSPG) 26, 36, 210 Hipl 2,5,6,109,110,127 Holoprosencephaly (HPE) 30, 41, 88, 148, 153-157, 161 HOX protein 14, 146

Idiopathic pulmonary fibrosis (IPF) 119-123, 125, 127 Immune system 70, 119, 120, 123 Indian hedgehog (IHH/Ihh) 35, 46, 87, 98, 101, 119, 120, 125, 126, 128, 146-151, 178,210,215-218,220,221 Injury 13,41, 119-126, 128, 188 Insulin-like growth factor (IGF) 107, 200 Interleukin-10 (IL-10) 123 Internalization 26, 27, 38, 127 Intestine 3,41, 120, 126

227

Index

K

N

Keratinocyte growth factor (KGF) 121-123, 125, 127

N-myc 12, 59, G6, 99, 107-109, 189-191, 194, 196-203 Neural stem cell (NSC) 3, 107, 188, 189, 202, 213, 222 Neurons 6, 12, 14, 59, 89, 90, 123, 127, 177-183, 187, 188, 192-195, 198, 212, 213,218 Nevoid basal cell carcinoma syndrome (NBCCS) 30, 53, 54, 56, 7A, 102 Noggin A7, 49 Non-small cell lung cancer (NSCLC) 123, 218 Nonsteroidal anti-inflammatory agents (NSAIDs) 68,69

Limb anomalies 129, 131, 158 Limb patterning 137,141 Lipid 25, 35-37, 77, 210, 222 Lung 4, 5, 13, 41, 42, 55, 7A, 83, 88-91, 94, 95, 101, 104, 107, 119-128,211, 215-218, 221 Lysosome 37, 39

M Malformations 2, 130, 131, 137, 139, 140, 142, 143, 147, 149, 150, 153, 155, 157, 158, 162, 167, 168, 180, 182, 184 Medulloblastoma (MB) 2, A-G, 8, 9, 11, 30, 42, 53, 54, 56-59, 65, 66, 71, 81, 101, 104, 107, 109, 110, 156, 158, 193, 195, 197, 199, 201-203, 210-213, 220, 222 Mice 1, 3-9, 11, 13, 14, 25, 35, 43-49, 53, 54, 56-60, 63-71, 74-76, 80-82, 86-95, 97-99, 102, 103, 105-107, 109, 110, 119-122, 124-126, 128, 131, 137-143, 147, 150, 151, 158, 159, 161, 162, 166, 167, 169, 170, 177, 178, 180-183, 191-199, 201, 202, 211, 212, 215, 216, 218-220 Mitogens 13, 86, 108, 188-191, 195, 197 Model system 24, 30, 123, 153, 155, 157-161, 166, 168 Modifiers 2, 55, 58, 131, 133 Morphogen 13, 24-26, 30, 34, 38, 41, 42, 45, 46, 49, 86, 120, 127, 128, 137, 146, 157, 160, 162, 164, 210 Mouse model 9, 63, 64, 71, 80, 102, 103, 105, 106, 110, 121, 122, 125, 131, 159, 201,212 Muscle 3, 13, 54, 65, 89, 101, 107, 127, 218 Mutations 1-5, 23, 26-30, 41, 46, 49, 53-56, 58-60, 63-66, 71, 74, 75, 78-80, 88-91, 102, 104-106, 110, 127, 129-134, 138-140, 142, 143, 146-151, 155, 156, 158, 159, 161, 168, 169, 181-183, 192, 195, 199, 210, 211, 215, 216, 218-221

o odontogenesis 169 Oligodactyly 137, 138 Organizing center 162, 168

Palatogenesis 168 Pallister-Hall syndrome (PHS) 3, 89, 130-134, 142, 143, 156 Pancreas 9, 13, 41, 101, 146, 151, 216-219, 221 Pancreatic adenocarcinoma 217, 219, 220 Papillogenesis 169 Patched 23, 7A, 75, 86, 102, 120, 126, 127, 141,148, 156, 177,193, 195, 210, 215, 216 Patched (PTCH) receptor 23, 148 Phenotype 1, 3, 23, 26, 30, 47-49, 75, 80, 88, 90-92, 98, 99, 101-106, 124, 126, 128-131, 133, 134, 138-143, 147, 149-151, 153, 155-157,159-161, 168, 180, 182-184, 188, 191, 192, 216, 220 Phosphorylation 28, 48, 59, 87, 189-192, 195, 198-203 PI-3 kinase 195,200-203 Polydactyly 30, 41, 53, 89, 90, 130-133, 137, 138, 141-143, 156, 159 pRb 191-193 see also Retinoblastoma protein Preaxial polydactylous (PPD) 138-140, 143 Progenitor cells 41-43, 45-47, 98, 101, 102, 105, 124, 125, 127, 128, 162, 183, 184, 200, 217-221


228 Prostate 4, 5, 8, 10-13, 41, 101, 104 Protein kinase A (PKA) 13, 48, 193, 203, 210 PTEN 203 Pulmonary fibrosis 119,127

R Receptor 23-28, 30, 34, 36, 50, 53, 59, 60, 70, 75, 86, 87, 107, 120, 123, 126, 128, 139,141, 148, 156, 160, 161, 168, 184, 187, 190, 193, 195, 200, 211, 212, 215, 216, 218, 222 Receptor tyrosine kinase (RTK) pathway 190, 195 Regeneration 12-14, 49, 120, 125, 127 Remodeling (A, 120, 125, 126, 190, 198 Repair 13, 14, 41, 42, 49, 101, 119-123, 125, 127, 128, 132, 188 Retinoblastoma protein 191 see also pRb Retinoid 69, 70, 160 Rhabdomyosarcoma (RMS) 9, 42, 53, 54, 56-60, 65, GG, 101

Signal transduction 23, 25-27, 37, 53, 74, 79, 80, 82, 86, 126, 127, 160, 161, 169, 178, 187,188, 194-196, 216, 218, 219 Skin 2-4, 6, 8, 9, 13, 30, 41- 43, 46, 48, 49, 53, 58, 63, 66-70, 7A, 75, 77. 80, 86, 88, 90, 92, 94, 95, 97-103, 105-110, 196, 210-212,218,222 Skin cancer 69, 86, 110 Smad 95, 128, 190, 203 Small cell lung cancer (SCLC) 5, 217, 218 Small molecule 6, 9, 13, 27, 34, 110, 211-213 Smoothened 2, 24, 53, 83, 86, 110, 120, 126, 127,141, 156, 159, 177, 184, 193, 195, 199, 202, 210, 211, 215, 216, 218, 220, 222 Sonic hedgehog (SHH/Shh) 1, 3, 4, 6-9, 12-14, 30, 34-39, 41-50, 53-55, 58-60, 65, G6. 7A-7G, 79-82, 86, 87, 91, 94-110, 119-128, 131, 132, 137-143, 146-148, 153,155-171, 177-184, 187, 189-203, 210-212,215-220,222 Stem cell 1, 3, 4, 6, 7, 13, 14, 77. 78, 86, 96, 97, 101, 102, 107, 125, 128, 187, 188, 202, 212, 213, 222

Sterol-sensing domain (SSD) 25, 27, 35 Stomach 5, 13, 120, 126, 216-218, 221 Submandibular gland (SMG) 170

Tcell receptor 123 Target genes 2, 27, 30, 43-50, 53, 65, 67. 68, 70, 75, 76. 80-83, 86, 87, 93-95, 97-99, 101, 102, 104-106, 108-110, 122, 141-143, 159, 178, 192, 198, 215, 216, 222 Telencephalon 91, 155, 164, 165, 168, 177-184, 218 Teratogens 157, 159-161 Transcription factor 2, 24, 47-49, 64, 74, 75, 79-83, 86, 87, 95, 109, 118, 120, 130, 139-141, 156, 162, 177, 178,189-194, 196, 200, 201, 203, 210, 215, 216, 222 Tumor suppressor 36, 53, 55, 57, 78, 126, 192,193,195,211,219,220 Tumors 1-6, 8-13, 30, 36, 41, 42, 45-50, 53-61, 63, 64-70, 74, 75, 78-83, 86, 87, 92, 96, 101-110, 126, 128, 156, 158, 177, 187, 191-193, 195, 196,199, 201-203, 210-213, 215, 218-222

Variant Clara cells (VCC) 124 Vioxx® 68

w Wing imaginal disc 23, 24 Wnt 13, 41, 45, ^7. 49, 80, 81, 100, 126, 159, 182-184, 190, 195,211,215, 220-222

Zone of polarizing activity (ZPA) 137-139, 141, 162, 178 ZPA regulatory sequence (ZRS) 139, 140

Hedgehog-Gli Signaling in Human Disease (Molecular Biology Intelligence Unit)

Ceramide Signaling (Molecular Biology Intelligence Unit)

Ubiquitin and Disease (Molecular Biology Intelligence Unit)

Endosomes (Molecular Biology Intelligence Unit)

Calreticulin (Molecular Biology Intelligence Unit)

p53 (Molecular Biology Intelligence Unit)

VEGF in Development (Molecular Biology Intelligence Unit)

Signaling in Human Disease MBIU

Molecular Biology of the Parathyroid (Molecular Biology Intelligence Unit)

Translation Mechanisms (Molecular Biology Intelligence Unit)

Cell-Cell Channels (Molecular Biology Intelligence Unit)

Lipocalins (Molecular Biology Intelligence Unit (Unnumbered).)

Protein Kinase C (Molecular Biology Intelligence Unit)

HOX Gene Expression (Molecular Biology Intelligence Unit)

Rho GTPases (Molecular Biology Intelligence Unit)

Tight Junctions (Molecular Biology Intelligence Unit)

Branching Morphogenesis (Molecular Biology Intelligence Unit)

Immunobiology of Carbohydrates (Molecular Biology Intelligence Unit)

The Nucleolus (Molecular Biology Intelligence Unit)

HOX Gene Expression (Molecular Biology Intelligence Unit)

Discovering Biomolecular Mechanisms with Computational Biology (Molecular Biology Intelligence Unit)

Molecular Basis Of Epithelial Appendage Morphogenesis (Molecular Biology Intelligence Unit)

Molecular Mechanisms of Smooth Muscle Contraction (Molecular Biology Intelligence Unit)

Discovering Biomolecular Mechanisms with Computational Biology (Molecular Biology Intelligence Unit)

Muscle Development in Drosophilia (Molecular Biology Intelligence Unit)

Behavioural And Morphological Asymmetries in Vertebrates (Molecular Biology Intelligence Unit)

Transmembrane Signaling Protocols (Methods in Molecular Biology)

Calcium Signaling Protocols (Methods in Molecular Biology)

Abl Family Kinases in Development and Disease (Molecular Biology Intelligence Unit)

DNA Repair and Human Disease (Medical Intelligence Unit)

The Jak-Stat Pathway in Hematopoiesis and Disease (Molecular Biology Intelligence Unit, 20)

Hedgehog-Gli Signaling in Human Disease (Molecular Biology Intelligence Unit)

Ceramide Signaling (Molecular Biology Intelligence Unit)

Ubiquitin and Disease (Molecular Biology Intelligence Unit)

Endosomes (Molecular Biology Intelligence Unit)

Calreticulin (Molecular Biology Intelligence Unit)

p53 (Molecular Biology Intelligence Unit)

VEGF in Development (Molecular Biology Intelligence Unit)

Signaling in Human Disease MBIU

Molecular Biology of the Parathyroid (Molecular Biology Intelligence Unit)

Translation Mechanisms (Molecular Biology Intelligence Unit)

Cell-Cell Channels (Molecular Biology Intelligence Unit)

Lipocalins (Molecular Biology Intelligence Unit (Unnumbered).)

Protein Kinase C (Molecular Biology Intelligence Unit)

HOX Gene Expression (Molecular Biology Intelligence Unit)

Rho GTPases (Molecular Biology Intelligence Unit)

Tight Junctions (Molecular Biology Intelligence Unit)

Branching Morphogenesis (Molecular Biology Intelligence Unit)

Immunobiology of Carbohydrates (Molecular Biology Intelligence Unit)

The Nucleolus (Molecular Biology Intelligence Unit)

HOX Gene Expression (Molecular Biology Intelligence Unit)

Discovering Biomolecular Mechanisms with Computational Biology (Molecular Biology Intelligence Unit)

Molecular Basis Of Epithelial Appendage Morphogenesis (Molecular Biology Intelligence Unit)

Molecular Mechanisms of Smooth Muscle Contraction (Molecular Biology Intelligence Unit)

Discovering Biomolecular Mechanisms with Computational Biology (Molecular Biology Intelligence Unit)

Muscle Development in Drosophilia (Molecular Biology Intelligence Unit)

Behavioural And Morphological Asymmetries in Vertebrates (Molecular Biology Intelligence Unit)

Transmembrane Signaling Protocols (Methods in Molecular Biology)

Calcium Signaling Protocols (Methods in Molecular Biology)

Abl Family Kinases in Development and Disease (Molecular Biology Intelligence Unit)

DNA Repair and Human Disease (Medical Intelligence Unit)

The Jak-Stat Pathway in Hematopoiesis and Disease (Molecular Biology Intelligence Unit, 20)

Recommend Documents